Re: [hlds] Post-outage thoughts
-- [ Picked text/plain from multipart/alternative ] It was a busy and windy week here at Valve. As most of the community knows by now, last week a major windstorm hit the greater Seattle area. Power outages are fairly common in the area, but the magnitude of this storm knocked out our datacenter, resulting in about 20 hours Steam interruption. This was very frustrating for everyone here and we know it was frustrating for you as well. We learned a lot and are taking steps to make sure that when the next storm like this hits our area 15-20 years from now *fingers crossed*, people will still be playing games on Steam while we are buying flashlight batteries. Looks like Valve are going to do something a bit more substantive than just protect their power supply. I could be wrong, but that is how I read this statement. On 12/22/06, Adam Sando [EMAIL PROTECTED] wrote: ARP/broadcast rules, especially when you have a customer whose network see's over 200mbps of it on a regular basis. Although pumping out 2Gbps out the wire to the internet does bring with it some networking overhead ;) The more hosts you have, and the bigger the subnets these machines live on, the more crazy ARP traffic you see. Who needs VLAN's these days anyway hehe ;) Regards, Adam. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Edward Luna Sent: Friday, 22 December 2006 3:20 AM To: hlds@list.valvesoftware.com Subject: RE: [hlds] Post-outage thoughts We may be saying the same thing. In a network comprised of 3 hubs (2 uplinks) all Ethernet traffic is offered to all ports on all hubs but on the same network using 3 switches, Ethernet traffic destined for a specific host (port) on switch 3 will only be presented to that port. Broadcasts are presented to the entire network in all cases. -Original Message- From: Whisper [mailto:[EMAIL PROTECTED] Sent: Thursday, December 21, 2006 9:31 AM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts -- [ Picked text/plain from multipart/alternative ] That is very bad The point of running a switch in the first place is to microsegment your network so every port becomes a collision domain. Where is that guy with then CCNA when you need him. Collisions are not the problem anyhow on switched Ethernet networks, it is broadcasts. On 12/22/06, Edward Luna [EMAIL PROTECTED] wrote: I agree that switches are a technical leap forward from hubs but to say hubs suck is to say networks sucked before switches were prevalent and that simply is not true. Although there are numerous differences between switches and hubs (especially managed switches) the most striking performance factor is that switches keep track of hosts relative to MAC address and discriminate between nodes while hubs present all Ethernet traffic to all hosts on the network. This feature of switches is essential in larger networks (say 48 hosts and up with heavy Ethernet traffic) in order to limit collisions, but of absolutely no consequence in a small network. With today's super smart switches, collisions may have been eliminated entirely... I'm not certain of that however, anyone who has managed an Ethernet network with over 48 hosts is well aware of the performance degradation caused by collisions in networks with hubs. Rule of thumb... large network use switches; small network, a hub will be fine (if you can even find one anymore hehehe). -Original Message- From: chad [mailto:[EMAIL PROTECTED] Sent: Wednesday, December 20, 2006 6:35 PM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts HUBS suck for more than 2 computers, and cost more than switches as you cannot get them new anymore at stores. however hubs are perfect for packet sniffing, and extending a cable past the recommended cable max length, other than that they are not economical, or sensible. that said I just got a hub for sniffing and extending cables if need be. is undetectable packet sniffing on switched networks easy (without managed switches) Hexis wrote: On Wed, Dec 20, 2006 at 08:53:20AM +, Gigabit Nick wrote: Most modern ADSL/Wireless routers have auto sensing non-manageable switches in them because the hardware is cheap and packet sniffing made people wary of hubs. Not so much. Hubs offer less performance due to their nature. At this point there is little or no advantage to a hub over a switch, and significant disabvantages. The market has migrated to small unmanaged switches being the norm for home networking. Now it will cost you more to buy a hub instead of a switch. Hubs have become speciality items for specific purposes. That and packet sniffing on a switched network is pretty trivial. Not as simple as on a hub, but still quite easy
RE: [hlds] Post-outage thoughts
I agree that switches are a technical leap forward from hubs but to say hubs suck is to say networks sucked before switches were prevalent and that simply is not true. Although there are numerous differences between switches and hubs (especially managed switches) the most striking performance factor is that switches keep track of hosts relative to MAC address and discriminate between nodes while hubs present all Ethernet traffic to all hosts on the network. This feature of switches is essential in larger networks (say 48 hosts and up with heavy Ethernet traffic) in order to limit collisions, but of absolutely no consequence in a small network. With today's super smart switches, collisions may have been eliminated entirely... I'm not certain of that however, anyone who has managed an Ethernet network with over 48 hosts is well aware of the performance degradation caused by collisions in networks with hubs. Rule of thumb... large network use switches; small network, a hub will be fine (if you can even find one anymore hehehe). -Original Message- From: chad [mailto:[EMAIL PROTECTED] Sent: Wednesday, December 20, 2006 6:35 PM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts HUBS suck for more than 2 computers, and cost more than switches as you cannot get them new anymore at stores. however hubs are perfect for packet sniffing, and extending a cable past the recommended cable max length, other than that they are not economical, or sensible. that said I just got a hub for sniffing and extending cables if need be. is undetectable packet sniffing on switched networks easy (without managed switches) Hexis wrote: On Wed, Dec 20, 2006 at 08:53:20AM +, Gigabit Nick wrote: Most modern ADSL/Wireless routers have auto sensing non-manageable switches in them because the hardware is cheap and packet sniffing made people wary of hubs. Not so much. Hubs offer less performance due to their nature. At this point there is little or no advantage to a hub over a switch, and significant disabvantages. The market has migrated to small unmanaged switches being the norm for home networking. Now it will cost you more to buy a hub instead of a switch. Hubs have become speciality items for specific purposes. That and packet sniffing on a switched network is pretty trivial. Not as simple as on a hub, but still quite easy. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Post-outage thoughts
-- [ Picked text/plain from multipart/alternative ] That is very bad The point of running a switch in the first place is to microsegment your network so every port becomes a collision domain. Where is that guy with then CCNA when you need him. Collisions are not the problem anyhow on switched Ethernet networks, it is broadcasts. On 12/22/06, Edward Luna [EMAIL PROTECTED] wrote: I agree that switches are a technical leap forward from hubs but to say hubs suck is to say networks sucked before switches were prevalent and that simply is not true. Although there are numerous differences between switches and hubs (especially managed switches) the most striking performance factor is that switches keep track of hosts relative to MAC address and discriminate between nodes while hubs present all Ethernet traffic to all hosts on the network. This feature of switches is essential in larger networks (say 48 hosts and up with heavy Ethernet traffic) in order to limit collisions, but of absolutely no consequence in a small network. With today's super smart switches, collisions may have been eliminated entirely... I'm not certain of that however, anyone who has managed an Ethernet network with over 48 hosts is well aware of the performance degradation caused by collisions in networks with hubs. Rule of thumb... large network use switches; small network, a hub will be fine (if you can even find one anymore hehehe). -Original Message- From: chad [mailto:[EMAIL PROTECTED] Sent: Wednesday, December 20, 2006 6:35 PM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts HUBS suck for more than 2 computers, and cost more than switches as you cannot get them new anymore at stores. however hubs are perfect for packet sniffing, and extending a cable past the recommended cable max length, other than that they are not economical, or sensible. that said I just got a hub for sniffing and extending cables if need be. is undetectable packet sniffing on switched networks easy (without managed switches) Hexis wrote: On Wed, Dec 20, 2006 at 08:53:20AM +, Gigabit Nick wrote: Most modern ADSL/Wireless routers have auto sensing non-manageable switches in them because the hardware is cheap and packet sniffing made people wary of hubs. Not so much. Hubs offer less performance due to their nature. At this point there is little or no advantage to a hub over a switch, and significant disabvantages. The market has migrated to small unmanaged switches being the norm for home networking. Now it will cost you more to buy a hub instead of a switch. Hubs have become speciality items for specific purposes. That and packet sniffing on a switched network is pretty trivial. Not as simple as on a hub, but still quite easy. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds -- ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: [hlds] Post-outage thoughts
-- [ Picked text/plain from multipart/alternative ] Most modern ADSL/Wireless routers have auto sensing non-manageable switches in them because the hardware is cheap and packet sniffing made people wary of hubs. So the guy could have a $70 wireless ADSL route and technically have a switch and router in his room. From: [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts Date: Tue, 19 Dec 2006 20:48:41 -0500 did somebody say they have switches and routers in thier dorm room on ups's? what in the world would you need a switch in your dorm room for? multiple vlans in your room? anyway, redundancy is always important. but you are right, it is not cost effective. Remember though, not that steam supports more and more games (and gets paid for it) would it not be more attractive for developers to know that someone will be able to play thier game no matter what the weather is like in seattle? - Original Message - From: Scott Tuttle [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Tuesday, December 19, 2006 8:07 PM Subject: RE: [hlds] Post-outage thoughts You would have to ask their sales force if being able to say the system is redundant would help them make sales.-Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of bob johnson Sent: Tuesday, December 19, 2006 5:45 PM To: hlds@list.valvesoftware.com Subject: RE: [hlds] Post-outage thoughts This is a multi-part message in MIME format. -- [ Picked text/plain from multipart/alternative ] Still sounds like alot of money invested for nothing. Where is the ROI for Valve? Will this cost them sales? I say no, the software has been bought and paid for. By the time the next release rolls around everyone will have forgotten about this. Did they lose data? Nope. Lawsuits? Nope. Why then would they dump many thousands of dollars into a more reliable and redundant system? Businesses do things that will help them make money and this won't. Anyone from Valve care to chime in?? -Original Message- From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED] On Behalf Of chad Sent: Tuesday, December 19, 2006 4:50 PM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts You are right, networking 101 teaches that, I took the networking class in high school, and got my NET+ certification, Even I know that backups are important, and I'm a college freshman (well I have enough credits to be a sophomore) I have backups, and they work, my laptop decided one day to disable all the user accounts and delete any that were useful (even administrator was disabled) i got back up and running in 5 minutes, and in 15 i had my programs back, and in the end spent more time trying to fix the laptop than trying to recover from it. after taking networking 101, I now have my computers (and switches and routers) in my dorm on UPSes, and on a cart that in a power outage of my floor could be wheeled downstairs, or in an Internet outage, I have a very long (100 M) cable to reach any other part of the building, and I could make it longer with a repeater if I had to, but it is long enough. Valve has a lotbigger budget, and paying customers and a lot better trained people than me, and should have much better backups and redundancy, I can survive a loss of service to my floor, with minimal downtime of my services, so valve should be able to stand the loss of a city (even by nuke) with minimal affect to the rest of the world, if the whole US was out of power, or gone by nuking, then I would be okay with them being off line, for a few days Scott Tuttle wrote: Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You think it looks good to investors that the backbone of the system went down for the entire world because of one geological disaster? You think that's a good selling point for software developers that want to bring their product to market? 273,468 game players couldn't play because Valve had all their eggs in that one geographical basket. Wise business decision? You decide... Ok maybe they are 500 level courses but you still get the point :D -Original Message- From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:57 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage
Re: [hlds] Post-outage thoughts
And what does this have to do with the outage? Alex Mottshaw wrote: Dear Alfred and Valve, First of all, let me say that I have had a lot of enjoyment from Counter Strike Source, so much so that I created GotGames.com.au in Australia to specifically address the lack of a serious CSS This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to [EMAIL PROTECTED] ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Post-outage thoughts
On Wed, Dec 20, 2006 at 08:53:20AM +, Gigabit Nick wrote: Most modern ADSL/Wireless routers have auto sensing non-manageable switches in them because the hardware is cheap and packet sniffing made people wary of hubs. Not so much. Hubs offer less performance due to their nature. At this point there is little or no advantage to a hub over a switch, and significant disabvantages. The market has migrated to small unmanaged switches being the norm for home networking. Now it will cost you more to buy a hub instead of a switch. Hubs have become speciality items for specific purposes. That and packet sniffing on a switched network is pretty trivial. Not as simple as on a hub, but still quite easy. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Post-outage thoughts
Possible the same reason as gotgames gets spammed across or forums by a couple of people, and usually promptly removed. Kingsley - Original Message - From: Steven Hartland [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Thursday, December 21, 2006 1:21 AM Subject: Re: [hlds] Post-outage thoughts And what does this have to do with the outage? Alex Mottshaw wrote: Dear Alfred and Valve, First of all, let me say that I have had a lot of enjoyment from Counter Strike Source, so much so that I created GotGames.com.au in Australia to specifically address the lack of a serious CSS This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to [EMAIL PROTECTED] ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Post-outage thoughts
2 swithces (soon to be 3) 2 wrt54g router-switch combos computers that act as routers. 0: Gigabit Nicks Gigabit Nick wrote: -- [ Picked text/plain from multipart/alternative ] Most modern ADSL/Wireless routers have auto sensing non-manageable switches in them because the hardware is cheap and packet sniffing made people wary of hubs. So the guy could have a $70 wireless ADSL route and technically have a switch and router in his room. From: [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts Date: Tue, 19 Dec 2006 20:48:41 -0500 did somebody say they have switches and routers in thier dorm room on ups's? what in the world would you need a switch in your dorm room for? multiple vlans in your room? anyway, redundancy is always important. but you are right, it is not cost effective. Remember though, not that steam supports more and more games (and gets paid for it) would it not be more attractive for developers to know that someone will be able to play thier game no matter what the weather is like in seattle? - Original Message - From: Scott Tuttle [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Tuesday, December 19, 2006 8:07 PM Subject: RE: [hlds] Post-outage thoughtsYou would have to ask their sales force if being able to say the system is redundant would help them make sales.-Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of bob johnson Sent: Tuesday, December 19, 2006 5:45 PM To: hlds@list.valvesoftware.com Subject: RE: [hlds] Post-outage thoughts This is a multi-part message in MIME format. -- [ Picked text/plain from multipart/alternative ] Still sounds like alot of money invested for nothing. Where is the ROI for Valve? Will this cost them sales? I say no, the software has been bought and paid for. By the time the next release rolls around everyone will have forgotten about this. Did they lose data? Nope. Lawsuits? Nope. Why then would they dump many thousands of dollars into a more reliable and redundant system? Businesses do things that will help them make money and this won't. Anyone from Valve care to chime in?? -Original Message- From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED] On Behalf Of chad Sent: Tuesday, December 19, 2006 4:50 PM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts You are right, networking 101 teaches that, I took the networking class in high school, and got my NET+ certification, Even I know that backups are important, and I'm a college freshman (well I have enough credits to be a sophomore) I have backups, and they work, my laptop decided one day to disable all the user accounts and delete any that were useful (even administrator was disabled) i got back up and running in 5 minutes, and in 15 i had my programs back, and in the end spent more time trying to fix the laptop than trying to recover from it. after taking networking 101, I now have my computers (and switches and routers) in my dorm on UPSes, and on a cart that in a power outage of my floor could be wheeled downstairs, or in an Internet outage, I have a very long (100 M) cable to reach any other part of the building, and I could make it longer with a repeater if I had to, but it is long enough. Valve has a lotbigger budget, and paying customers and a lot better trained people than me, and should have much better backups and redundancy, I can survive a loss of service to my floor, with minimal downtime of my services, so valve should be able to stand the loss of a city (even by nuke) with minimal affect to the rest of the world, if the whole US was out of power, or gone by nuking, then I would be okay with them being off line, for a few days Scott Tuttle wrote: Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You think it looks good to investors that the backbone of the system went down for the entire world because of one geological disaster? You think that's a good selling point for software developers that want to bring their product to market? 273,468 game players couldn't play because Valve had all their eggs in that one geographical basket. Wise business decision? You decide... Ok maybe they are 500 level courses but you still get the point :D-Original Message- From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:57 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage
RE: RE: RE: Re: [hlds] Post-outage thoughts
Whether or not a service provider chooses to deploy redundant services is a decision that is generally made as part of an overall risk-management analysis. Factors such as probability of component failure, business impact and cost are weighed in reaching a decision as to how much money a provider should (and can afford) to invest in redundant service elements. While a systemic power outage is a possibility, it may not be very probable. In fact, there is every likelihood that service elements which would be affected by such a wide outage are not all within Valve's control. We have no information regarding Valve's service infrastructure, but we might assume that it includes fault-tolerant elements (e.g. clustered servers, redundant network paths, etc.) which have been chosen to provide protection from more probable outages (for example, individual hardware failures, network outage of a given carrier). Given the funding resources to do so, most service providers would eagerly embrace geographic redundancy. However, no business has unlimited financial resources and in the end, Valve has to strike a balance between cost and risk, in delivering its services. Valve has an obligation to its investors to make balanced spending decisions and deliver sustainable profitability as much as it needs to deliver reasonable service levels to its customers. As well, the cost of complete redundancy would almost certainly have to be borne in the price of the product. While the end-user impact was certainly real, it is not, after all, an air traffic control system. last night, our servers were full again. I think Valve did a respectable job in restoring services in a timely fashion. No doubt they were extremely motivated to do so. It appeared to me that they followed a prioritized approach, first restoring services critical to supporting game-play. While this simply may have been a sequence imposed by the situation, versus any kind of altruistic service policy, the net effect was the same. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Scott Tuttle Sent: Monday, December 18, 2006 4:23 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You think it looks good to investors that the backbone of the system went down for the entire world because of one geological disaster? You think that's a good selling point for software developers that want to bring their product to market? 273,468 game players couldn't play because Valve had all their eggs in that one geographical basket. Wise business decision? You decide... Ok maybe they are 500 level courses but you still get the point :D -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:57 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts All I'm seeing is whining, pettiness, and monday morning quarterbacking. Lets try this. If anyone out there has a diagram of the Valve infrastructure, and a complete understanding of who they contract with for what services and facilities, then lets see it. I only am reading people bitching about what Valve should have done over the last 10 years, and I could do it better, without any reguard or perspective on what the real world impact things may be having in the Seattle area. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: RE: RE: Re: [hlds] Post-outage thoughts
Very well said Frazer, as always. However, I'm obligated to point out, whatever fault tolerance Valve may or may not have built in... it was insufficient for this event. Until we are informed to the contrary by Valve, we must conclude that they were not geographically redundant... furthermore, to assume they considered a wide-spread power outage in the Northwest not very probable does not bode well for their level of fault tolerance analysis. We needn't wonder if their plan would work, we know it failed. The salient question to be answered now is do they intend to bring their redundancy inline with the need and if not... will their customers accept that position? -Original Message- From: Frazer [mailto:[EMAIL PROTECTED] Sent: Tuesday, December 19, 2006 7:43 AM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Whether or not a service provider chooses to deploy redundant services is a decision that is generally made as part of an overall risk-management analysis. Factors such as probability of component failure, business impact and cost are weighed in reaching a decision as to how much money a provider should (and can afford) to invest in redundant service elements. While a systemic power outage is a possibility, it may not be very probable. In fact, there is every likelihood that service elements which would be affected by such a wide outage are not all within Valve's control. We have no information regarding Valve's service infrastructure, but we might assume that it includes fault-tolerant elements (e.g. clustered servers, redundant network paths, etc.) which have been chosen to provide protection from more probable outages (for example, individual hardware failures, network outage of a given carrier). Given the funding resources to do so, most service providers would eagerly embrace geographic redundancy. However, no business has unlimited financial resources and in the end, Valve has to strike a balance between cost and risk, in delivering its services. Valve has an obligation to its investors to make balanced spending decisions and deliver sustainable profitability as much as it needs to deliver reasonable service levels to its customers. As well, the cost of complete redundancy would almost certainly have to be borne in the price of the product. While the end-user impact was certainly real, it is not, after all, an air traffic control system. last night, our servers were full again. I think Valve did a respectable job in restoring services in a timely fashion. No doubt they were extremely motivated to do so. It appeared to me that they followed a prioritized approach, first restoring services critical to supporting game-play. While this simply may have been a sequence imposed by the situation, versus any kind of altruistic service policy, the net effect was the same. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Scott Tuttle Sent: Monday, December 18, 2006 4:23 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You think it looks good to investors that the backbone of the system went down for the entire world because of one geological disaster? You think that's a good selling point for software developers that want to bring their product to market? 273,468 game players couldn't play because Valve had all their eggs in that one geographical basket. Wise business decision? You decide... Ok maybe they are 500 level courses but you still get the point :D -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:57 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts All I'm seeing is whining, pettiness, and monday morning quarterbacking. Lets try this. If anyone out there has a diagram of the Valve infrastructure, and a complete understanding of who they contract with for what services and facilities, then lets see it. I only am reading people bitching about what Valve should have done over the last 10 years, and I could do it better, without any reguard or perspective on what the real world impact things may be having in the Seattle area. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Post-outage thoughts
All of these post on this subject and still NOTHING FROM VALVE!! Any bets on what their gonna do? My moneys on nothing From: Edward Luna [EMAIL PROTECTED] Date: 2006/12/19 Tue AM 07:18:14 CST To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Very well said Frazer, as always. However, I'm obligated to point out, whatever fault tolerance Valve may or may not have built in... it was insufficient for this event. Until we are informed to the contrary by Valve, we must conclude that they were not geographically redundant... furthermore, to assume they considered a wide-spread power outage in the Northwest not very probable does not bode well for their level of fault tolerance analysis. We needn't wonder if their plan would work, we know it failed. The salient question to be answered now is do they intend to bring their redundancy inline with the need and if not... will their customers accept that position? -Original Message- From: Frazer [mailto:[EMAIL PROTECTED] Sent: Tuesday, December 19, 2006 7:43 AM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Whether or not a service provider chooses to deploy redundant services is a decision that is generally made as part of an overall risk-management analysis. Factors such as probability of component failure, business impact and cost are weighed in reaching a decision as to how much money a provider should (and can afford) to invest in redundant service elements. While a systemic power outage is a possibility, it may not be very probable. In fact, there is every likelihood that service elements which would be affected by such a wide outage are not all within Valve's control. We have no information regarding Valve's service infrastructure, but we might assume that it includes fault-tolerant elements (e.g. clustered servers, redundant network paths, etc.) which have been chosen to provide protection from more probable outages (for example, individual hardware failures, network outage of a given carrier). Given the funding resources to do so, most service providers would eagerly embrace geographic redundancy. However, no business has unlimited financial resources and in the end, Valve has to strike a balance between cost and risk, in delivering its services. Valve has an obligation to its investors to make balanced spending decisions and deliver sustainable profitability as much as it needs to deliver reasonable service levels to its customers. As well, the cost of complete redundancy would almost certainly have to be borne in the price of the product. While the end-user impact was certainly real, it is not, after all, an air traffic control system. last night, our servers were full again. I think Valve did a respectable job in restoring services in a timely fashion. No doubt they were extremely motivated to do so. It appeared to me that they followed a prioritized approach, first restoring services critical to supporting game-play. While this simply may have been a sequence imposed by the situation, versus any kind of altruistic service policy, the net effect was the same. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Scott Tuttle Sent: Monday, December 18, 2006 4:23 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You think it looks good to investors that the backbone of the system went down for the entire world because of one geological disaster? You think that's a good selling point for software developers that want to bring their product to market? 273,468 game players couldn't play because Valve had all their eggs in that one geographical basket. Wise business decision? You decide... Ok maybe they are 500 level courses but you still get the point :D -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:57 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts All I'm seeing is whining, pettiness, and monday morning quarterbacking. Lets try this. If anyone out there has a diagram of the Valve infrastructure, and a complete understanding of who they contract with for what services and facilities, then lets see it. I only am reading people bitching about what Valve should have done over the last 10 years, and I could do it better, without any reguard or perspective on what the real world impact things may be having in the Seattle area
Re: [hlds] Post-outage thoughts
-- [ Picked text/plain from multipart/alternative ] One thing constantly being missed is that section C of paragraph 9 of Steam Subscriber agreement which every one of us agreed to states that: VALVE DOES NOT GUARANTEE CONTINUOUS, ERROR-FREE, VIRUS-FREE OR SECURE OPERATION AND ACCESS TO STEAM, THE STEAM SOFTWARE, YOUR ACCOUNT AND/OR YOUR SUBSCRIPTIONS(S). It means we all agreed with the fact that we can not demand Valve to support Steam at all. The fact that Valve restored the service reasonably quick means they don't want to loose customers and profit but does not mean they had obligations towards us to do so. Another thing that should be considered is overall network downtime throughout the year. What was that? less than 12 hours overall? Meaning availability is about 99.8%... Not the best figure for mission critical application but pretty much reasonable for gaming services. Regards, Newbie -Original Message- From: [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Date: Tue, 19 Dec 2006 7:36:13 -0600 Subject: Re: [hlds] Post-outage thoughts All of these post on this subject and still NOTHING FROM VALVE!! Any bets on what their gonna do? My moneys on nothing From: Edward Luna [EMAIL PROTECTED] Date: 2006/12/19 Tue AM 07:18:14 CST To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Very well said Frazer, as always. However, I'm obligated to point out, whatever fault tolerance Valve may or may not have built in... it was insufficient for this event. Until we are informed to the contrary by Valve, we must conclude that they were not geographically redundant... furthermore, to assume they considered a wide-spread power outage in the Northwest not very probable does not bode well for their level of fault tolerance analysis. We needn't wonder if their plan would work, we know it failed. The salient question to be answered now is do they intend to bring their redundancy inline with the need and if not... will their customers accept that position? -Original Message- From: Frazer [mailto:[EMAIL PROTECTED] Sent: Tuesday, December 19, 2006 7:43 AM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Whether or not a service provider chooses to deploy redundant services is a decision that is generally made as part of an overall risk-management analysis. Factors such as probability of component failure, business impact and cost are weighed in reaching a decision as to how much money a provider should (and can afford) to invest in redundant service elements. While a systemic power outage is a possibility, it may not be very probable. In fact, there is every likelihood that service elements which would be affected by such a wide outage are not all within Valve's control. We have no information regarding Valve's service infrastructure, but we might assume that it includes fault-tolerant elements (e.g. clustered servers, redundant network paths, etc.) which have been chosen to provide protection from more probable outages (for example, individual hardware failures, network outage of a given carrier). Given the funding resources to do so, most service providers would eagerly embrace geographic redundancy. However, no business has unlimited financial resources and in the end, Valve has to strike a balance between cost and risk, in delivering its services. Valve has an obligation to its investors to make balanced spending decisions and deliver sustainable profitability as much as it needs to deliver reasonable service levels to its customers. As well, the cost of complete redundancy would almost certainly have to be borne in the price of the product. While the end-user impact was certainly real, it is not, after all, an air traffic control system. last night, our servers were full again. I think Valve did a respectable job in restoring services in a timely fashion. No doubt they were extremely motivated to do so. It appeared to me that they followed a prioritized approach, first restoring services critical to supporting game-play. While this simply may have been a sequence imposed by the situation, versus any kind of altruistic service policy, the net effect was the same. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Scott Tuttle Sent: Monday, December 18, 2006 4:23 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You think it looks good to investors that the backbone of the system went down for the entire world
Re: [hlds] Post-outage thoughts
-- [ Picked text/plain from multipart/alternative ] Considering that STEAM is now a revenue generating service, my bet is Valve will sort this out sooner rather than later, especially since they are now responsible to not only their own games anymore but to a lot of other Game Developers as well. On 12/20/06, Newbie [EMAIL PROTECTED] wrote: -- [ Picked text/plain from multipart/alternative ] One thing constantly being missed is that section C of paragraph 9 of Steam Subscriber agreement which every one of us agreed to states that: VALVE DOES NOT GUARANTEE CONTINUOUS, ERROR-FREE, VIRUS-FREE OR SECURE OPERATION AND ACCESS TO STEAM, THE STEAM SOFTWARE, YOUR ACCOUNT AND/OR YOUR SUBSCRIPTIONS(S). It means we all agreed with the fact that we can not demand Valve to support Steam at all. The fact that Valve restored the service reasonably quick means they don't want to loose customers and profit but does not mean they had obligations towards us to do so. Another thing that should be considered is overall network downtime throughout the year. What was that? less than 12 hours overall? Meaning availability is about 99.8%... Not the best figure for mission critical application but pretty much reasonable for gaming services. Regards, Newbie -Original Message- From: [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Date: Tue, 19 Dec 2006 7:36:13 -0600 Subject: Re: [hlds] Post-outage thoughts All of these post on this subject and still NOTHING FROM VALVE!! Any bets on what their gonna do? My moneys on nothing From: Edward Luna [EMAIL PROTECTED] Date: 2006/12/19 Tue AM 07:18:14 CST To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Very well said Frazer, as always. However, I'm obligated to point out, whatever fault tolerance Valve may or may not have built in... it was insufficient for this event. Until we are informed to the contrary by Valve, we must conclude that they were not geographically redundant... furthermore, to assume they considered a wide-spread power outage in the Northwest not very probable does not bode well for their level of fault tolerance analysis. We needn't wonder if their plan would work, we know it failed. The salient question to be answered now is do they intend to bring their redundancy inline with the need and if not... will their customers accept that position? -Original Message- From: Frazer [mailto:[EMAIL PROTECTED] Sent: Tuesday, December 19, 2006 7:43 AM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Whether or not a service provider chooses to deploy redundant services is a decision that is generally made as part of an overall risk-management analysis. Factors such as probability of component failure, business impact and cost are weighed in reaching a decision as to how much money a provider should (and can afford) to invest in redundant service elements. While a systemic power outage is a possibility, it may not be very probable. In fact, there is every likelihood that service elements which would be affected by such a wide outage are not all within Valve's control. We have no information regarding Valve's service infrastructure, but we might assume that it includes fault-tolerant elements (e.g. clustered servers, redundant network paths, etc.) which have been chosen to provide protection from more probable outages (for example, individual hardware failures, network outage of a given carrier). Given the funding resources to do so, most service providers would eagerly embrace geographic redundancy. However, no business has unlimited financial resources and in the end, Valve has to strike a balance between cost and risk, in delivering its services. Valve has an obligation to its investors to make balanced spending decisions and deliver sustainable profitability as much as it needs to deliver reasonable service levels to its customers. As well, the cost of complete redundancy would almost certainly have to be borne in the price of the product. While the end-user impact was certainly real, it is not, after all, an air traffic control system. last night, our servers were full again. I think Valve did a respectable job in restoring services in a timely fashion. No doubt they were extremely motivated to do so. It appeared to me that they followed a prioritized approach, first restoring services critical to supporting game-play. While this simply may have been a sequence imposed by the situation, versus any kind of altruistic service policy, the net effect was the same. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Scott Tuttle Sent: Monday, December 18, 2006 4:23 PM To: hlds@list.valvesoftware.com Subject
RE: RE: Re: [hlds] Post-outage thoughts
Who was bitching about not being able to play CS anyways? I`m sorry for the people wherever they are all of them suffering anything I really am. I really don`t watch tv or news much. As for the topic of discussion with is HLDS ADMIN what exactly are u looking to hear from us so you can stop wasting our time? The discussion was about Valves obvious lack of re-investment in their network infrastructure and obviously qualified network support staff. Not about a snow storm or suffering people. Merry Christmas -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:43 PM To: hlds@list.valvesoftware.com Subject: RE: RE: Re: [hlds] Post-outage thoughts Then I guess I may have things backwards in thinking that any personal inconvenience should be outweighed by actual tragety suffered by others. Yeah, not oficially winter, so those people aren't oficially freezing. http://seattletimes.nwsource.com/html/localnews/2003482605_webstormmainb ar17.html Or am I also wrong in also thinking that maybe local repair efforts should be aimed at getting people heat, instead of making sure CS players can get online. And no matter what Valve did or didn't do, they are still relying on outside resources as much as anyone else in the area. Even the fact that it may be the Valve employees, or others that are in charge of keeping things running, are too busy keeping their families warm wouldn't be reason enough for some. Maybe we even found out why people feel that CS, DoD, et. al. are loosing players. It may possibly be the lack of soul and community. Last I checked Winter hasn't begun yet. :)~ Obviously people are not complaining about loss of service as being equivalent to loss of life or livelihood... I think you may have gone a bit overboard with that one. The majority of complaints are centered around an apparent lack of re-investment in infrastructure on Valve's part and it is a valid complaint. Perhaps we have demonstrated too much faith in Valves understanding of the most basic concept in network management... that being, no single point failure should bring a network down. That fact stands on it's own and need not be measured against a lost Christmas for those unfortunate to have been effected by the storms. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Sent: Monday, December 18, 2006 12:27 PM To: hlds@list.valvesoftware.com Subject: RE: Re: [hlds] Post-outage thoughts Lets keep things in perspective people. Over a million people without power or heat int the middle of winter. Christmas planning for lots of those are out the window because of lost wages, loss of life, and loss of basic services, and people are whining about not being able to play a computer game for a few days, and then exorcising Valve about not having things back online immediately. Lets think about things based on the effects on real people lives instead of using a tragedy to puff one's resume. It's nice to see the spirit of the holidays on display. Netcom had a massive outage in '96 that lasted almost the same duration as this valve outage. At 11:06 AM 12/18/2006, -Mike- wrote: There are far too many 90% empty datacenters practically sitting on top of major exchanges down here in the SF Bay Area (and all over the US) for Valve to have suffered the outage they saw due to storm conditions. I'm sorry, but a decent distributed network architecture with properly configured load balancing hardware takes care of these single points of failure. But hey, what do I know... I only managed Yahoo's mailservers at GlobalCenter, FriendFinder and Lycos' hardware at Exodus, built and managed bulletproof network backbones at @Home and Netcom... So it's not like I'd know anything about engineering a method of preventing a little lack of power, IP dialtone, or overload from taking your biggest cash machine offline. (sigh) Sorry Valve, I'm gainfully employed and I do not consult on the side. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: Roman Hatsiev [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Sunday, December 17, 2006 11:21:39 PM Subject: Re: [hlds] Post-outage thoughts This is true only as long as you work with tested and widely adopted solution like Active Directory. For closed proprietary system of Steam size designed without redundancy in mind this can be a kind of tricky exercise... Regards, Roman ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list
Re: [hlds] Post-outage thoughts
That would be nice but I think it would be even better for someone from Valve to say something. Yes, no, go piss up a rope. it don't matter as long as it gives an indication of what they think about this and what they are looking into doing (if anything) to correct it. From: Whisper [EMAIL PROTECTED] Date: 2006/12/19 Tue AM 10:45:00 CST To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts -- [ Picked text/plain from multipart/alternative ] Considering that STEAM is now a revenue generating service, my bet is Valve will sort this out sooner rather than later, especially since they are now responsible to not only their own games anymore but to a lot of other Game Developers as well. On 12/20/06, Newbie [EMAIL PROTECTED] wrote: -- [ Picked text/plain from multipart/alternative ] One thing constantly being missed is that section C of paragraph 9 of Steam Subscriber agreement which every one of us agreed to states that: VALVE DOES NOT GUARANTEE CONTINUOUS, ERROR-FREE, VIRUS-FREE OR SECURE OPERATION AND ACCESS TO STEAM, THE STEAM SOFTWARE, YOUR ACCOUNT AND/OR YOUR SUBSCRIPTIONS(S). It means we all agreed with the fact that we can not demand Valve to support Steam at all. The fact that Valve restored the service reasonably quick means they don't want to loose customers and profit but does not mean they had obligations towards us to do so. Another thing that should be considered is overall network downtime throughout the year. What was that? less than 12 hours overall? Meaning availability is about 99.8%... Not the best figure for mission critical application but pretty much reasonable for gaming services. Regards, Newbie -Original Message- From: [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Date: Tue, 19 Dec 2006 7:36:13 -0600 Subject: Re: [hlds] Post-outage thoughts All of these post on this subject and still NOTHING FROM VALVE!! Any bets on what their gonna do? My moneys on nothing From: Edward Luna [EMAIL PROTECTED] Date: 2006/12/19 Tue AM 07:18:14 CST To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Very well said Frazer, as always. However, I'm obligated to point out, whatever fault tolerance Valve may or may not have built in... it was insufficient for this event. Until we are informed to the contrary by Valve, we must conclude that they were not geographically redundant... furthermore, to assume they considered a wide-spread power outage in the Northwest not very probable does not bode well for their level of fault tolerance analysis. We needn't wonder if their plan would work, we know it failed. The salient question to be answered now is do they intend to bring their redundancy inline with the need and if not... will their customers accept that position? -Original Message- From: Frazer [mailto:[EMAIL PROTECTED] Sent: Tuesday, December 19, 2006 7:43 AM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Whether or not a service provider chooses to deploy redundant services is a decision that is generally made as part of an overall risk-management analysis. Factors such as probability of component failure, business impact and cost are weighed in reaching a decision as to how much money a provider should (and can afford) to invest in redundant service elements. While a systemic power outage is a possibility, it may not be very probable. In fact, there is every likelihood that service elements which would be affected by such a wide outage are not all within Valve's control. We have no information regarding Valve's service infrastructure, but we might assume that it includes fault-tolerant elements (e.g. clustered servers, redundant network paths, etc.) which have been chosen to provide protection from more probable outages (for example, individual hardware failures, network outage of a given carrier). Given the funding resources to do so, most service providers would eagerly embrace geographic redundancy. However, no business has unlimited financial resources and in the end, Valve has to strike a balance between cost and risk, in delivering its services. Valve has an obligation to its investors to make balanced spending decisions and deliver sustainable profitability as much as it needs to deliver reasonable service levels to its customers. As well, the cost of complete redundancy would almost certainly have to be borne in the price of the product. While the end-user impact was certainly real, it is not, after all, an air traffic control system. last night, our servers were full again. I think Valve did a respectable job
Re: [hlds] Post-outage thoughts
You are right, networking 101 teaches that, I took the networking class in high school, and got my NET+ certification, Even I know that backups are important, and I'm a college freshman (well I have enough credits to be a sophomore) I have backups, and they work, my laptop decided one day to disable all the user accounts and delete any that were useful (even administrator was disabled) i got back up and running in 5 minutes, and in 15 i had my programs back, and in the end spent more time trying to fix the laptop than trying to recover from it. after taking networking 101, I now have my computers (and switches and routers) in my dorm on UPSes, and on a cart that in a power outage of my floor could be wheeled downstairs, or in an Internet outage, I have a very long (100 M) cable to reach any other part of the building, and I could make it longer with a repeater if I had to, but it is long enough. Valve has a lotbigger budget, and paying customers and a lot better trained people than me, and should have much better backups and redundancy, I can survive a loss of service to my floor, with minimal downtime of my services, so valve should be able to stand the loss of a city (even by nuke) with minimal affect to the rest of the world, if the whole US was out of power, or gone by nuking, then I would be okay with them being off line, for a few days Scott Tuttle wrote: Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You think it looks good to investors that the backbone of the system went down for the entire world because of one geological disaster? You think that's a good selling point for software developers that want to bring their product to market? 273,468 game players couldn't play because Valve had all their eggs in that one geographical basket. Wise business decision? You decide... Ok maybe they are 500 level courses but you still get the point :D -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:57 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts All I'm seeing is whining, pettiness, and monday morning quarterbacking. Lets try this. If anyone out there has a diagram of the Valve infrastructure, and a complete understanding of who they contract with for what services and facilities, then lets see it. I only am reading people bitching about what Valve should have done over the last 10 years, and I could do it better, without any reguard or perspective on what the real world impact things may be having in the Seattle area. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: [hlds] Post-outage thoughts
This is a multi-part message in MIME format. -- [ Picked text/plain from multipart/alternative ] Still sounds like alot of money invested for nothing. Where is the ROI for Valve? Will this cost them sales? I say no, the software has been bought and paid for. By the time the next release rolls around everyone will have forgotten about this. Did they lose data? Nope. Lawsuits? Nope. Why then would they dump many thousands of dollars into a more reliable and redundant system? Businesses do things that will help them make money and this won't. Anyone from Valve care to chime in?? -Original Message- From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED] On Behalf Of chad Sent: Tuesday, December 19, 2006 4:50 PM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts You are right, networking 101 teaches that, I took the networking class in high school, and got my NET+ certification, Even I know that backups are important, and I'm a college freshman (well I have enough credits to be a sophomore) I have backups, and they work, my laptop decided one day to disable all the user accounts and delete any that were useful (even administrator was disabled) i got back up and running in 5 minutes, and in 15 i had my programs back, and in the end spent more time trying to fix the laptop than trying to recover from it. after taking networking 101, I now have my computers (and switches and routers) in my dorm on UPSes, and on a cart that in a power outage of my floor could be wheeled downstairs, or in an Internet outage, I have a very long (100 M) cable to reach any other part of the building, and I could make it longer with a repeater if I had to, but it is long enough. Valve has a lotbigger budget, and paying customers and a lot better trained people than me, and should have much better backups and redundancy, I can survive a loss of service to my floor, with minimal downtime of my services, so valve should be able to stand the loss of a city (even by nuke) with minimal affect to the rest of the world, if the whole US was out of power, or gone by nuking, then I would be okay with them being off line, for a few days Scott Tuttle wrote: Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You think it looks good to investors that the backbone of the system went down for the entire world because of one geological disaster? You think that's a good selling point for software developers that want to bring their product to market? 273,468 game players couldn't play because Valve had all their eggs in that one geographical basket. Wise business decision? You decide... Ok maybe they are 500 level courses but you still get the point :D -Original Message- From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:57 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts All I'm seeing is whining, pettiness, and monday morning quarterbacking. Lets try this. If anyone out there has a diagram of the Valve infrastructure, and a complete understanding of who they contract with for what services and facilities, then lets see it. I only am reading people bitching about what Valve should have done over the last 10 years, and I could do it better, without any reguard or perspective on what the real world impact things may be having in the Seattle area. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds -- Cal, Stanford -Students Only Classifieds Students only , Pay nothing to Buy Sell textbooks, furniture more. http://a8-asy.a8ww.net/a8-ads/adftrclick?redirectid=20e9b7562a9c927234e3ca61ecb4660b -- ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: [hlds] Post-outage thoughts
You would have to ask their sales force if being able to say the system is redundant would help them make sales. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of bob johnson Sent: Tuesday, December 19, 2006 5:45 PM To: hlds@list.valvesoftware.com Subject: RE: [hlds] Post-outage thoughts This is a multi-part message in MIME format. -- [ Picked text/plain from multipart/alternative ] Still sounds like alot of money invested for nothing. Where is the ROI for Valve? Will this cost them sales? I say no, the software has been bought and paid for. By the time the next release rolls around everyone will have forgotten about this. Did they lose data? Nope. Lawsuits? Nope. Why then would they dump many thousands of dollars into a more reliable and redundant system? Businesses do things that will help them make money and this won't. Anyone from Valve care to chime in?? -Original Message- From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED] On Behalf Of chad Sent: Tuesday, December 19, 2006 4:50 PM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts You are right, networking 101 teaches that, I took the networking class in high school, and got my NET+ certification, Even I know that backups are important, and I'm a college freshman (well I have enough credits to be a sophomore) I have backups, and they work, my laptop decided one day to disable all the user accounts and delete any that were useful (even administrator was disabled) i got back up and running in 5 minutes, and in 15 i had my programs back, and in the end spent more time trying to fix the laptop than trying to recover from it. after taking networking 101, I now have my computers (and switches and routers) in my dorm on UPSes, and on a cart that in a power outage of my floor could be wheeled downstairs, or in an Internet outage, I have a very long (100 M) cable to reach any other part of the building, and I could make it longer with a repeater if I had to, but it is long enough. Valve has a lotbigger budget, and paying customers and a lot better trained people than me, and should have much better backups and redundancy, I can survive a loss of service to my floor, with minimal downtime of my services, so valve should be able to stand the loss of a city (even by nuke) with minimal affect to the rest of the world, if the whole US was out of power, or gone by nuking, then I would be okay with them being off line, for a few days Scott Tuttle wrote: Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You think it looks good to investors that the backbone of the system went down for the entire world because of one geological disaster? You think that's a good selling point for software developers that want to bring their product to market? 273,468 game players couldn't play because Valve had all their eggs in that one geographical basket. Wise business decision? You decide... Ok maybe they are 500 level courses but you still get the point :D -Original Message- From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:57 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts All I'm seeing is whining, pettiness, and monday morning quarterbacking. Lets try this. If anyone out there has a diagram of the Valve infrastructure, and a complete understanding of who they contract with for what services and facilities, then lets see it. I only am reading people bitching about what Valve should have done over the last 10 years, and I could do it better, without any reguard or perspective on what the real world impact things may be having in the Seattle area. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds -- Cal, Stanford -Students Only Classifieds Students only , Pay nothing to Buy Sell textbooks, furniture more. http://a8-asy.a8ww.net/a8-ads/adftrclick?redirectid=20e9b7562a 9c927234e3ca61ecb4660b -- ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: [hlds] Post-outage thoughts
Dear Alfred and Valve, First of all, let me say that I have had a lot of enjoyment from Counter Strike Source, so much so that I created GotGames.com.au in Australia to specifically address the lack of a serious CSS competitive community in Australia and New Zealand. In a little over 6 months, we have created the biggest source community in Australia and New Zealand. The updates and changes that you have made to the game have been very positive on the whole. Fixing the crouch POV and the SourceTV was a huge boost for the competitive scene and much appreciated. I am writing to you in the hope that you will address 4 issues that several programmers believe would take one competent programmer less than 1 day resolve. If these 4 simple changes were made, it would really help the competitive community and help organisations such as WCG, CPL and CEVO make the change to source. These issues involve no changes to the engine itself but merely the addition of some simple cvars. 1. Dead body Cam, when you die you have approximately 3-5 seconds to tell your teammates through Ventrilo and Teamspeak which way the opposition went before the camera view changes to one of your teammates. A cvar that turns off the dead body cam so that when you die, the camera view instantly changes to your teammates POV removing the ability for you to be able to spectate the enemy illegally while you are dead. This simple cvar would make the world of difference to the online competitive community. 2. Dead players being able to tell their live teammates through Ventrilo or Teamspeak that the opposition has just picked up the bomb because the scoreboard tells players when a bomb has been picked up the opposition. A simple cvar would resolve this and could be implemented very easily. 3. The creation of a cvar that turns the need to purchase ammo on, obviously this would be more involved than the first 2 but couldn't be to hard assuming that you still have the ammo code somewhere. By removing the need to purchase ammo you removed a substantial amount of tactics from the game, I also agree though removing ammo for the average player is a good idea, hence the best option is a cvar. 4. A cvar that increases the walking speed for competitive purposes, again a very simple change that would be welcomed by the competitive community without affecting the public community. So as you can see, these are very simple and easy changes that Valve could make that would significantly improve the competitive community and I'm certain that 95% of the competitive community would agree with me. Alex Hybrid Mottshaw This message is intended solely for the individual (s) and entity(s) addressed. It is confidential and may contain legally privileged information. The use, copying or distribution of this message or any information it contains, by anyone other than the addressee, is prohibited. If you have received this message in error, please notify [EMAIL PROTECTED] * -Original Message- From: Whisper [mailto:[EMAIL PROTECTED] Sent: Wednesday, 20 December 2006 3:45 AM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts -- [ Picked text/plain from multipart/alternative ] Considering that STEAM is now a revenue generating service, my bet is Valve will sort this out sooner rather than later, especially since they are now responsible to not only their own games anymore but to a lot of other Game Developers as well. On 12/20/06, Newbie [EMAIL PROTECTED] wrote: -- [ Picked text/plain from multipart/alternative ] One thing constantly being missed is that section C of paragraph 9 of Steam Subscriber agreement which every one of us agreed to states that: VALVE DOES NOT GUARANTEE CONTINUOUS, ERROR-FREE, VIRUS-FREE OR SECURE OPERATION AND ACCESS TO STEAM, THE STEAM SOFTWARE, YOUR ACCOUNT AND/OR YOUR SUBSCRIPTIONS(S). It means we all agreed with the fact that we can not demand Valve to support Steam at all. The fact that Valve restored the service reasonably quick means they don't want to loose customers and profit but does not mean they had obligations towards us to do so. Another thing that should be considered is overall network downtime throughout the year. What was that? less than 12 hours overall? Meaning availability is about 99.8%... Not the best figure for mission critical application but pretty much reasonable for gaming services. Regards, Newbie -Original Message- From: [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Date: Tue, 19 Dec 2006 7:36:13 -0600 Subject: Re: [hlds] Post-outage thoughts All of these post on this subject and still NOTHING FROM VALVE!! Any bets on what their gonna do? My moneys on nothing From: Edward Luna [EMAIL PROTECTED] Date: 2006/12/19 Tue AM 07:18:14 CST To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts Very well said Frazer, as always. However, I'm obligated
Re: [hlds] Post-outage thoughts
did somebody say they have switches and routers in thier dorm room on ups's? what in the world would you need a switch in your dorm room for? multiple vlans in your room? anyway, redundancy is always important. but you are right, it is not cost effective. Remember though, not that steam supports more and more games (and gets paid for it) would it not be more attractive for developers to know that someone will be able to play thier game no matter what the weather is like in seattle? - Original Message - From: Scott Tuttle [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Tuesday, December 19, 2006 8:07 PM Subject: RE: [hlds] Post-outage thoughts You would have to ask their sales force if being able to say the system is redundant would help them make sales. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of bob johnson Sent: Tuesday, December 19, 2006 5:45 PM To: hlds@list.valvesoftware.com Subject: RE: [hlds] Post-outage thoughts This is a multi-part message in MIME format. -- [ Picked text/plain from multipart/alternative ] Still sounds like alot of money invested for nothing. Where is the ROI for Valve? Will this cost them sales? I say no, the software has been bought and paid for. By the time the next release rolls around everyone will have forgotten about this. Did they lose data? Nope. Lawsuits? Nope. Why then would they dump many thousands of dollars into a more reliable and redundant system? Businesses do things that will help them make money and this won't. Anyone from Valve care to chime in?? -Original Message- From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED] On Behalf Of chad Sent: Tuesday, December 19, 2006 4:50 PM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts You are right, networking 101 teaches that, I took the networking class in high school, and got my NET+ certification, Even I know that backups are important, and I'm a college freshman (well I have enough credits to be a sophomore) I have backups, and they work, my laptop decided one day to disable all the user accounts and delete any that were useful (even administrator was disabled) i got back up and running in 5 minutes, and in 15 i had my programs back, and in the end spent more time trying to fix the laptop than trying to recover from it. after taking networking 101, I now have my computers (and switches and routers) in my dorm on UPSes, and on a cart that in a power outage of my floor could be wheeled downstairs, or in an Internet outage, I have a very long (100 M) cable to reach any other part of the building, and I could make it longer with a repeater if I had to, but it is long enough. Valve has a lotbigger budget, and paying customers and a lot better trained people than me, and should have much better backups and redundancy, I can survive a loss of service to my floor, with minimal downtime of my services, so valve should be able to stand the loss of a city (even by nuke) with minimal affect to the rest of the world, if the whole US was out of power, or gone by nuking, then I would be okay with them being off line, for a few days Scott Tuttle wrote: Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You think it looks good to investors that the backbone of the system went down for the entire world because of one geological disaster? You think that's a good selling point for software developers that want to bring their product to market? 273,468 game players couldn't play because Valve had all their eggs in that one geographical basket. Wise business decision? You decide... Ok maybe they are 500 level courses but you still get the point :D -Original Message- From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:57 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts All I'm seeing is whining, pettiness, and monday morning quarterbacking. Lets try this. If anyone out there has a diagram of the Valve infrastructure, and a complete understanding of who they contract with for what services and facilities, then lets see it. I only am reading people bitching about what Valve should have done over the last 10 years, and I could do it better, without any reguard or perspective on what the real world impact things may be having in the Seattle area
Re: [hlds] Post-outage thoughts
I have switches to connect my domain controller to my file server to my laptop. the UPS was $1 at the school surplus sale, and was not even my best find. How else will I back up my profile automatically, and install software on my HP laptop that breaks like every 3 months and needs to get software and data back on it again. Also, I use the switch to connect my game servers (dual pIII 1.3 ghz with soon to be a gig of ram each, and no CD drive, or usb booting support that cost $5.00 each and come with a 4 hour parts delivery warranty until October next year) to my RIS server to install windows without CD's, but with cheap cd keys we get at school. If you know a better way for free to get access to my 950 gigabyte file server, while it is not a DC because that slows it down a ton, and have the ability to access it securely from anywhere on the internet, install windows without a cd drive, and be able to instantly recover from my laptop deciding it doesn't like me, please email me at [EMAIL PROTECTED] oh and soon I will be doing off site replication of important shares between my dorm and my parents house in case of a disaster like the sprinklers spraying that rusty copper conducting water on my servers, or their house burning down. no vlans here, switches are like $2.00 for a 16 port 10/100 with more than 1.6 gbps internal bandwidth NIB on ebay, so I just use switches and wrt54g's with dd-wrt v.23 firmware to segment my network. oh and yes I did say that. Ryan Brady wrote: did somebody say they have switches and routers in thier dorm room on ups's? what in the world would you need a switch in your dorm room for? multiple vlans in your room? anyway, redundancy is always important. but you are right, it is not cost effective. Remember though, not that steam supports more and more games (and gets paid for it) would it not be more attractive for developers to know that someone will be able to play thier game no matter what the weather is like in seattle? - Original Message - From: Scott Tuttle [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Tuesday, December 19, 2006 8:07 PM Subject: RE: [hlds] Post-outage thoughts You would have to ask their sales force if being able to say the system is redundant would help them make sales. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of bob johnson Sent: Tuesday, December 19, 2006 5:45 PM To: hlds@list.valvesoftware.com Subject: RE: [hlds] Post-outage thoughts This is a multi-part message in MIME format. -- [ Picked text/plain from multipart/alternative ] Still sounds like alot of money invested for nothing. Where is the ROI for Valve? Will this cost them sales? I say no, the software has been bought and paid for. By the time the next release rolls around everyone will have forgotten about this. Did they lose data? Nope. Lawsuits? Nope. Why then would they dump many thousands of dollars into a more reliable and redundant system? Businesses do things that will help them make money and this won't. Anyone from Valve care to chime in?? -Original Message- From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED] On Behalf Of chad Sent: Tuesday, December 19, 2006 4:50 PM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts You are right, networking 101 teaches that, I took the networking class in high school, and got my NET+ certification, Even I know that backups are important, and I'm a college freshman (well I have enough credits to be a sophomore) I have backups, and they work, my laptop decided one day to disable all the user accounts and delete any that were useful (even administrator was disabled) i got back up and running in 5 minutes, and in 15 i had my programs back, and in the end spent more time trying to fix the laptop than trying to recover from it. after taking networking 101, I now have my computers (and switches and routers) in my dorm on UPSes, and on a cart that in a power outage of my floor could be wheeled downstairs, or in an Internet outage, I have a very long (100 M) cable to reach any other part of the building, and I could make it longer with a repeater if I had to, but it is long enough. Valve has a lotbigger budget, and paying customers and a lot better trained people than me, and should have much better backups and redundancy, I can survive a loss of service to my floor, with minimal downtime of my services, so valve should be able to stand the loss of a city (even by nuke) with minimal affect to the rest of the world, if the whole US was out of power, or gone by nuking, then I would be okay with them being off line, for a few days Scott Tuttle wrote: Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You
Re: [hlds] Post-outage thoughts
There are far too many 90% empty datacenters practically sitting on top of major exchanges down here in the SF Bay Area (and all over the US) for Valve to have suffered the outage they saw due to storm conditions. I'm sorry, but a decent distributed network architecture with properly configured load balancing hardware takes care of these single points of failure. But hey, what do I know... I only managed Yahoo's mailservers at GlobalCenter, FriendFinder and Lycos' hardware at Exodus, built and managed bulletproof network backbones at @Home and Netcom... So it's not like I'd know anything about engineering a method of preventing a little lack of power, IP dialtone, or overload from taking your biggest cash machine offline. (sigh) Sorry Valve, I'm gainfully employed and I do not consult on the side. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: Roman Hatsiev [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Sunday, December 17, 2006 11:21:39 PM Subject: Re: [hlds] Post-outage thoughts This is true only as long as you work with tested and widely adopted solution like Active Directory. For closed proprietary system of Steam size designed without redundancy in mind this can be a kind of tricky exercise... Regards, Roman ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Post-outage thoughts
Netcom had a massive outage in '96 that lasted almost the same duration as this valve outage. At 11:06 AM 12/18/2006, -Mike- wrote: There are far too many 90% empty datacenters practically sitting on top of major exchanges down here in the SF Bay Area (and all over the US) for Valve to have suffered the outage they saw due to storm conditions. I'm sorry, but a decent distributed network architecture with properly configured load balancing hardware takes care of these single points of failure. But hey, what do I know... I only managed Yahoo's mailservers at GlobalCenter, FriendFinder and Lycos' hardware at Exodus, built and managed bulletproof network backbones at @Home and Netcom... So it's not like I'd know anything about engineering a method of preventing a little lack of power, IP dialtone, or overload from taking your biggest cash machine offline. (sigh) Sorry Valve, I'm gainfully employed and I do not consult on the side. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: Roman Hatsiev [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Sunday, December 17, 2006 11:21:39 PM Subject: Re: [hlds] Post-outage thoughts This is true only as long as you work with tested and widely adopted solution like Active Directory. For closed proprietary system of Steam size designed without redundancy in mind this can be a kind of tricky exercise... Regards, Roman ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Post-outage thoughts
And what'd we learn from that... 1 ~ When Rieger hands you the company AMEX and said go get generators because the mezzanine datacenter is on UPS for about two more hours before the entire shell/UUCP/NNTP network goes dark... go to a commercial generator company, not Home Depot. 2 ~ Reiger will toss a chair through a plate glass window and Indiana Jones 30 feet down an extension cord to the ground if it means plugging in to a running generator(s) in the back of a pickup truck with 10 minutes to spare on the UPS banks. 3 ~ You can run an an entire Internet Service Provider on gasoline for an extended period of time, but entropy is still a bitch... and will get you in the end. 4 ~ Learn from your mistakes, and do not repeat them. Reiger likes to yell. Netcom had one major outage. One. I was glad I was there for it, it was a memorable experience. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: Gary Stanley [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Monday, December 18, 2006 8:36:02 AM Subject: Re: [hlds] Post-outage thoughts Netcom had a massive outage in '96 that lasted almost the same duration as this valve outage. At 11:06 AM 12/18/2006, -Mike- wrote: There are far too many 90% empty datacenters practically sitting on top of major exchanges down here in the SF Bay Area (and all over the US) for Valve to have suffered the outage they saw due to storm conditions. I'm sorry, but a decent distributed network architecture with properly configured load balancing hardware takes care of these single points of failure. But hey, what do I know... I only managed Yahoo's mailservers at GlobalCenter, FriendFinder and Lycos' hardware at Exodus, built and managed bulletproof network backbones at @Home and Netcom... So it's not like I'd know anything about engineering a method of preventing a little lack of power, IP dialtone, or overload from taking your biggest cash machine offline. (sigh) Sorry Valve, I'm gainfully employed and I do not consult on the side. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: Roman Hatsiev [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Sunday, December 17, 2006 11:21:39 PM Subject: Re: [hlds] Post-outage thoughts This is true only as long as you work with tested and widely adopted solution like Active Directory. For closed proprietary system of Steam size designed without redundancy in mind this can be a kind of tricky exercise... Regards, Roman ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Post-outage thoughts
Sorry but a DC without full UPS and generator backed power is just NOT a DC its a bedroom / shed operation. As others have pointed out running such an operation from a single DC is beyond belief as well. I do hope heads roll for this as its caused major pain. Steve This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to [EMAIL PROTECTED] ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Post-outage thoughts
-- [ Picked text/plain from multipart/alternative ] Why are you guys even worrying about power?? I know it was a power issue that caused the outage, but the significant fact to be discovered in all of this, is that there is this great big stinking single point of failure. All it needs is somebody to come along with a backhoe, or to do some arc welding, and it won't matter how much you have backed up your power then. The fact of the matter is, that the entire STEAM system needs to be distributed geographically so no matter what the weather, the power, the Internet conditions, or whatever of the 1001 things that can go wrong that could go wrong, more to the point, it is the things you don't think of that tend to get you, at least Valve will have covered their bases as best as humanely possible. On 12/19/06, Steven Hartland [EMAIL PROTECTED] wrote: Sorry but a DC without full UPS and generator backed power is just NOT a DC its a bedroom / shed operation. As others have pointed out running such an operation from a single DC is beyond belief as well. I do hope heads roll for this as its caused major pain. Steve This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to [EMAIL PROTECTED] ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds -- ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: Re: [hlds] Post-outage thoughts
Lets keep things in perspective people. Over a million people without power or heat int the middle of winter. Christmas planning for lots of those are out the window because of lost wages, loss of life, and loss of basic services, and people are whining about not being able to play a computer game for a few days, and then exorcising Valve about not having things back online immediately. Lets think about things based on the effects on real people lives instead of using a tragedy to puff one's resume. It's nice to see the spirit of the holidays on display. Netcom had a massive outage in '96 that lasted almost the same duration as this valve outage. At 11:06 AM 12/18/2006, -Mike- wrote: There are far too many 90% empty datacenters practically sitting on top of major exchanges down here in the SF Bay Area (and all over the US) for Valve to have suffered the outage they saw due to storm conditions. I'm sorry, but a decent distributed network architecture with properly configured load balancing hardware takes care of these single points of failure. But hey, what do I know... I only managed Yahoo's mailservers at GlobalCenter, FriendFinder and Lycos' hardware at Exodus, built and managed bulletproof network backbones at @Home and Netcom... So it's not like I'd know anything about engineering a method of preventing a little lack of power, IP dialtone, or overload from taking your biggest cash machine offline. (sigh) Sorry Valve, I'm gainfully employed and I do not consult on the side. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: Roman Hatsiev [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Sunday, December 17, 2006 11:21:39 PM Subject: Re: [hlds] Post-outage thoughts This is true only as long as you work with tested and widely adopted solution like Active Directory. For closed proprietary system of Steam size designed without redundancy in mind this can be a kind of tricky exercise... Regards, Roman ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: Re: [hlds] Post-outage thoughts
Last I checked Winter hasn't begun yet. :)~ Obviously people are not complaining about loss of service as being equivalent to loss of life or livelihood... I think you may have gone a bit overboard with that one. The majority of complaints are centered around an apparent lack of re-investment in infrastructure on Valve's part and it is a valid complaint. Perhaps we have demonstrated too much faith in Valves understanding of the most basic concept in network management... that being, no single point failure should bring a network down. That fact stands on it's own and need not be measured against a lost Christmas for those unfortunate to have been effected by the storms. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Sent: Monday, December 18, 2006 12:27 PM To: hlds@list.valvesoftware.com Subject: RE: Re: [hlds] Post-outage thoughts Lets keep things in perspective people. Over a million people without power or heat int the middle of winter. Christmas planning for lots of those are out the window because of lost wages, loss of life, and loss of basic services, and people are whining about not being able to play a computer game for a few days, and then exorcising Valve about not having things back online immediately. Lets think about things based on the effects on real people lives instead of using a tragedy to puff one's resume. It's nice to see the spirit of the holidays on display. Netcom had a massive outage in '96 that lasted almost the same duration as this valve outage. At 11:06 AM 12/18/2006, -Mike- wrote: There are far too many 90% empty datacenters practically sitting on top of major exchanges down here in the SF Bay Area (and all over the US) for Valve to have suffered the outage they saw due to storm conditions. I'm sorry, but a decent distributed network architecture with properly configured load balancing hardware takes care of these single points of failure. But hey, what do I know... I only managed Yahoo's mailservers at GlobalCenter, FriendFinder and Lycos' hardware at Exodus, built and managed bulletproof network backbones at @Home and Netcom... So it's not like I'd know anything about engineering a method of preventing a little lack of power, IP dialtone, or overload from taking your biggest cash machine offline. (sigh) Sorry Valve, I'm gainfully employed and I do not consult on the side. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: Roman Hatsiev [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Sunday, December 17, 2006 11:21:39 PM Subject: Re: [hlds] Post-outage thoughts This is true only as long as you work with tested and widely adopted solution like Active Directory. For closed proprietary system of Steam size designed without redundancy in mind this can be a kind of tricky exercise... Regards, Roman ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: Re: Re: [hlds] Post-outage thoughts
If a power outage that affected over a million poeple in a large geographical area (and is STILL affecting people) isn't a good enough explanation for a short outage, then I'm thinking there may not be one good enough for some around here. I'm reasonably sure that if it was a small yeild nuke, we'd be hearing about how valve servers should have been rad-hardened. Valve was not the only people to go down. Banking, commerce, and other private and public networks also were affected. No company can predict and prepare for every systemic failure that can happen. And yes, I am and administrator. I have been for around 5 years. The only thing that valve OWES me is a REASONABLE attempt to keep things running. And to all the admins out there without hearts three sizes too small. Merry Christmas, and Happy New Year! Honestly, this is an administrators list. Most of us here are going to very coldly and to the fact discuss the impact of the outage. If you are overly sensitive or feel the need for a sleeve to cry on about the human side of the Northwest storm then I'd suggest you unsubscribe and head over to Fark for a group hug. Valve should provide us with a reasonable explanation as to why they have all of their eggs in one geographical basket. They've had nearly ten years to plan and deploy preventative measures against such a failure. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: [EMAIL PROTECTED] [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Monday, December 18, 2006 9:26:59 AM Subject: RE: Re: [hlds] Post-outage thoughts Lets keep things in perspective people. Over a million people without power or heat int the middle of winter. Christmas planning for lots of those are out the window because of lost wages, loss of life, and loss of basic services, and people are whining about not being able to play a computer game for a few days, and then exorcising Valve about not having things back online immediately. Lets think about things based on the effects on real people lives instead of using a tragedy to puff one's resume. It's nice to see the spirit of the holidays on display. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: Re: Re: [hlds] Post-outage thoughts
The point that you seem to be missing is that, considering the size of the Steam network and the sheer number of users, there should have been more than one location for the Steam network to fall back on. Many of the companies you mentioned probably were rather small in comparison; any large company should have a sense of redundancy and backup. What if a tree fell on that building? That would have been the end of Valve, because they couldn't recover from losing their entire network like that. Yes, people are suffering from a storm, but people have to suffer through horrid issues all the time. The troubles of one geographic location should not affect another one halfway around the world. Geographic redundancy would have prevented this issue. --- [EMAIL PROTECTED] wrote: If a power outage that affected over a million poeple in a large geographical area (and is STILL affecting people) isn't a good enough explanation for a short outage, then I'm thinking there may not be one good enough for some around here. I'm reasonably sure that if it was a small yeild nuke, we'd be hearing about how valve servers should have been rad-hardened. Valve was not the only people to go down. Banking, commerce, and other private and public networks also were affected. No company can predict and prepare for every systemic failure that can happen. And yes, I am and administrator. I have been for around 5 years. The only thing that valve OWES me is a REASONABLE attempt to keep things running. And to all the admins out there without hearts three sizes too small. Merry Christmas, and Happy New Year! Honestly, this is an administrators list. Most of us here are going to very coldly and to the fact discuss the impact of the outage. If you are overly sensitive or feel the need for a sleeve to cry on about the human side of the Northwest storm then I'd suggest you unsubscribe and head over to Fark for a group hug. Valve should provide us with a reasonable explanation as to why they have all of their eggs in one geographical basket. They've had nearly ten years to plan and deploy preventative measures against such a failure. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: [EMAIL PROTECTED] [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Monday, December 18, 2006 9:26:59 AM Subject: RE: Re: [hlds] Post-outage thoughts Lets keep things in perspective people. Over a million people without power or heat int the middle of winter. Christmas planning for lots of those are out the window because of lost wages, loss of life, and loss of basic services, and people are whining about not being able to play a computer game for a few days, and then exorcising Valve about not having things back online immediately. Lets think about things based on the effects on real people lives instead of using a tragedy to puff one's resume. It's nice to see the spirit of the holidays on display. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: Re: [hlds] Post-outage thoughts
An explanation from Valve? Don't hold your breath. The only one I have seen explain anything is Alfred. They probably rag on him big time for wasting his time on this list instead of working on projects that actually make Valve money. We won't matter to them unless a money man says so From: -Mike- [EMAIL PROTECTED] Date: 2006/12/18 Mon PM 12:36:32 CST To: hlds@list.valvesoftware.com Subject: Re: Re: [hlds] Post-outage thoughts Honestly, this is an administrators list. Most of us here are going to very coldly and to the fact discuss the impact of the outage. If you are overly sensitive or feel the need for a sleeve to cry on about the human side of the Northwest storm then I'd suggest you unsubscribe and head over to Fark for a group hug. Valve should provide us with a reasonable explanation as to why they have all of their eggs in one geographical basket. They've had nearly ten years to plan and deploy preventative measures against such a failure. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: [EMAIL PROTECTED] [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Monday, December 18, 2006 9:26:59 AM Subject: RE: Re: [hlds] Post-outage thoughts Lets keep things in perspective people. Over a million people without power or heat int the middle of winter. Christmas planning for lots of those are out the window because of lost wages, loss of life, and loss of basic services, and people are whining about not being able to play a computer game for a few days, and then exorcising Valve about not having things back online immediately. Lets think about things based on the effects on real people lives instead of using a tragedy to puff one's resume. It's nice to see the spirit of the holidays on display. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: RE: Re: [hlds] Post-outage thoughts
... You are missing the point. Yes, people are suffering. We get that. It's unfortunate, but it is unavoidable; a natural disaster struck and you just gotta get through it. Valve's issue, on the other hand, WAS avoidable. A redundant network would have prevented this whole mess. When did we ever condemn those who were suffering? You just brought in another issue for no reason, as it has no effect on a gaming network. You make it sound like I am responsible for them suffering while I am 2000+ miles away.. --- [EMAIL PROTECTED] wrote: Then I guess I may have things backwards in thinking that any personal inconvenience should be outweighed by actual tragety suffered by others. Yeah, not oficially winter, so those people aren't oficially freezing. http://seattletimes.nwsource.com/html/localnews/2003482605_webstormmainbar17.html Or am I also wrong in also thinking that maybe local repair efforts should be aimed at getting people heat, instead of making sure CS players can get online. And no matter what Valve did or didn't do, they are still relying on outside resources as much as anyone else in the area. Even the fact that it may be the Valve employees, or others that are in charge of keeping things running, are too busy keeping their families warm wouldn't be reason enough for some. Maybe we even found out why people feel that CS, DoD, et. al. are loosing players. It may possibly be the lack of soul and community. Last I checked Winter hasn't begun yet. :)~ Obviously people are not complaining about loss of service as being equivalent to loss of life or livelihood... I think you may have gone a bit overboard with that one. The majority of complaints are centered around an apparent lack of re-investment in infrastructure on Valve's part and it is a valid complaint. Perhaps we have demonstrated too much faith in Valves understanding of the most basic concept in network management... that being, no single point failure should bring a network down. That fact stands on it's own and need not be measured against a lost Christmas for those unfortunate to have been effected by the storms. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Sent: Monday, December 18, 2006 12:27 PM To: hlds@list.valvesoftware.com Subject: RE: Re: [hlds] Post-outage thoughts Lets keep things in perspective people. Over a million people without power or heat int the middle of winter. Christmas planning for lots of those are out the window because of lost wages, loss of life, and loss of basic services, and people are whining about not being able to play a computer game for a few days, and then exorcising Valve about not having things back online immediately. Lets think about things based on the effects on real people lives instead of using a tragedy to puff one's resume. It's nice to see the spirit of the holidays on display. Netcom had a massive outage in '96 that lasted almost the same duration as this valve outage. At 11:06 AM 12/18/2006, -Mike- wrote: There are far too many 90% empty datacenters practically sitting on top of major exchanges down here in the SF Bay Area (and all over the US) for Valve to have suffered the outage they saw due to storm conditions. I'm sorry, but a decent distributed network architecture with properly configured load balancing hardware takes care of these single points of failure. But hey, what do I know... I only managed Yahoo's mailservers at GlobalCenter, FriendFinder and Lycos' hardware at Exodus, built and managed bulletproof network backbones at @Home and Netcom... So it's not like I'd know anything about engineering a method of preventing a little lack of power, IP dialtone, or overload from taking your biggest cash machine offline. (sigh) Sorry Valve, I'm gainfully employed and I do not consult on the side. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: Roman Hatsiev [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Sunday, December 17, 2006 11:21:39 PM Subject: Re: [hlds] Post-outage thoughts This is true only as long as you work with tested and widely adopted solution like Active Directory. For closed proprietary system of Steam size designed without redundancy in mind this can be a kind of tricky exercise... Regards, Roman ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please
RE: RE: Re: [hlds] Post-outage thoughts
Then I guess I may have things backwards in thinking that any personal inconvenience should be outweighed by actual tragety suffered by others. Yeah, not oficially winter, so those people aren't oficially freezing. http://seattletimes.nwsource.com/html/localnews/2003482605_webstormmainbar17.html Or am I also wrong in also thinking that maybe local repair efforts should be aimed at getting people heat, instead of making sure CS players can get online. And no matter what Valve did or didn't do, they are still relying on outside resources as much as anyone else in the area. Even the fact that it may be the Valve employees, or others that are in charge of keeping things running, are too busy keeping their families warm wouldn't be reason enough for some. Maybe we even found out why people feel that CS, DoD, et. al. are loosing players. It may possibly be the lack of soul and community. Last I checked Winter hasn't begun yet. :)~ Obviously people are not complaining about loss of service as being equivalent to loss of life or livelihood... I think you may have gone a bit overboard with that one. The majority of complaints are centered around an apparent lack of re-investment in infrastructure on Valve's part and it is a valid complaint. Perhaps we have demonstrated too much faith in Valves understanding of the most basic concept in network management... that being, no single point failure should bring a network down. That fact stands on it's own and need not be measured against a lost Christmas for those unfortunate to have been effected by the storms. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Sent: Monday, December 18, 2006 12:27 PM To: hlds@list.valvesoftware.com Subject: RE: Re: [hlds] Post-outage thoughts Lets keep things in perspective people. Over a million people without power or heat int the middle of winter. Christmas planning for lots of those are out the window because of lost wages, loss of life, and loss of basic services, and people are whining about not being able to play a computer game for a few days, and then exorcising Valve about not having things back online immediately. Lets think about things based on the effects on real people lives instead of using a tragedy to puff one's resume. It's nice to see the spirit of the holidays on display. Netcom had a massive outage in '96 that lasted almost the same duration as this valve outage. At 11:06 AM 12/18/2006, -Mike- wrote: There are far too many 90% empty datacenters practically sitting on top of major exchanges down here in the SF Bay Area (and all over the US) for Valve to have suffered the outage they saw due to storm conditions. I'm sorry, but a decent distributed network architecture with properly configured load balancing hardware takes care of these single points of failure. But hey, what do I know... I only managed Yahoo's mailservers at GlobalCenter, FriendFinder and Lycos' hardware at Exodus, built and managed bulletproof network backbones at @Home and Netcom... So it's not like I'd know anything about engineering a method of preventing a little lack of power, IP dialtone, or overload from taking your biggest cash machine offline. (sigh) Sorry Valve, I'm gainfully employed and I do not consult on the side. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: Roman Hatsiev [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Sunday, December 17, 2006 11:21:39 PM Subject: Re: [hlds] Post-outage thoughts This is true only as long as you work with tested and widely adopted solution like Active Directory. For closed proprietary system of Steam size designed without redundancy in mind this can be a kind of tricky exercise... Regards, Roman ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: RE: RE: Re: [hlds] Post-outage thoughts
Steam network went down. Steam network did not need to go down. Therefore, something could have been done to prevent the Steam network from going down. That's all there is to it. We are not discussing real-world impact because this is not a list meant for that kind of discussion. This is a discussion about the Steam server platform. The network went down and players were unable to play. Why should we come up with a new infrastructure for Valve? They have 100 or so employees, they should do it themselves; it should have been done a long time ago. THAT IS THE POINT. --- [EMAIL PROTECTED] wrote: All I'm seeing is whining, pettiness, and monday morning quarterbacking. Lets try this. If anyone out there has a diagram of the Valve infrastructure, and a complete understanding of who they contract with for what services and facilities, then lets see it. I only am reading people bitching about what Valve should have done over the last 10 years, and I could do it better, without any reguard or perspective on what the real world impact things may be having in the Seattle area. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: RE: RE: Re: [hlds] Post-outage thoughts
Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You think it looks good to investors that the backbone of the system went down for the entire world because of one geological disaster? You think that's a good selling point for software developers that want to bring their product to market? 273,468 game players couldn't play because Valve had all their eggs in that one geographical basket. Wise business decision? You decide... Ok maybe they are 500 level courses but you still get the point :D -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:57 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts All I'm seeing is whining, pettiness, and monday morning quarterbacking. Lets try this. If anyone out there has a diagram of the Valve infrastructure, and a complete understanding of who they contract with for what services and facilities, then lets see it. I only am reading people bitching about what Valve should have done over the last 10 years, and I could do it better, without any reguard or perspective on what the real world impact things may be having in the Seattle area. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Post-outage thoughts
atouk, none of us are happy about the effects of the power outage, however as with any natural disaster there are consequences that can be mitigated and consequences that can be reacted to. valve now have a secure stable network of content servers that managed to keep working perfectly throughout the power outage, we are all questioning why they dont have a similar multi homed network of authentication servers, and in this context you are doing nothing but trolling. On Monday 18 December 2006 19:57, [EMAIL PROTECTED] wrote: All I'm seeing is whining, pettiness, and monday morning quarterbacking. Lets try this. If anyone out there has a diagram of the Valve infrastructure, and a complete understanding of who they contract with for what services and facilities, then lets see it. I only am reading people bitching about what Valve should have done over the last 10 years, and I could do it better, without any reguard or perspective on what the real world impact things may be having in the Seattle area. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: RE: RE: RE: Re: [hlds] Post-outage thoughts
Steam went down - Absolutely Steam did not need to go down. - Possibly Something could have been done. - Probobly My argument isn't with Steam going down, Steam not going down, but with attitudes here. It's three days later, people are STILL without power up there, and yet we have people complaining, demanding explanations, and flashing credentials as if they know step by step and circuit by circuit what happened and why. On Long Island last year there was a power outage that lasted for over a week, and the report on the cause took almost 6 months without finding a definite cause. And that was without the area wide problems that are happening up there. The to hell with the situation, I demand you fix it NOW and tell me WHY it went down attitude, while root causes of the problem are still being worked on attitudes more than slightly disturbing. None of my posts have been in defense of Valve per. se. Steam network went down. Steam network did not need to go down. Therefore, something could have been done to prevent the Steam network from going down. That's all there is to it. We are not discussing real-world impact because this is not a list meant for that kind of discussion. This is a discussion about the Steam server platform. The network went down and players were unable to play. Why should we come up with a new infrastructure for Valve? They have 100 or so employees, they should do it themselves; it should have been done a long time ago. THAT IS THE POINT. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: RE: RE: RE: Re: [hlds] Post-outage thoughts
Ok, last post unless there is a specific point to counter, and a change of tact. I'd like to congratulate everyone at Valve, and especially the people in the IT department for restoring services after a major and widespread weather related disaster. Well done! Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You think it looks good to investors that the backbone of the system went down for the entire world because of one geological disaster? You think that's a good selling point for software developers that want to bring their product to market? 273,468 game players couldn't play because Valve had all their eggs in that one geographical basket. Wise business decision? You decide... Ok maybe they are 500 level courses but you still get the point :D -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:57 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts All I'm seeing is whining, pettiness, and monday morning quarterbacking. Lets try this. If anyone out there has a diagram of the Valve infrastructure, and a complete understanding of who they contract with for what services and facilities, then lets see it. I only am reading people bitching about what Valve should have done over the last 10 years, and I could do it better, without any reguard or perspective on what the real world impact things may be having in the Seattle area. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: RE: Re: [hlds] Post-outage thoughts
I'm sorry my friend... I think you are missing the point entirely. There is not a single person on this list who would want his access to a CS server restored before essential services were restored to people in need due to a sever storm... that's not the issue. Of course we want real human tragedies addressed before entertainment... we are not questioning the order in which services are restored. We are talking about a properly designed network with appropriate redundancy designed in so that a single point failure does not bring down the entire network to begin with. Of course, after the tragedy hit, all efforts should have been on helping people. The broader point is that the network should have continued to function worldwide with only local outages. Valves network is obviously not designed properly... they should fix it, as soon as all the humanitarian issues are addressed. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:43 PM To: hlds@list.valvesoftware.com Subject: RE: RE: Re: [hlds] Post-outage thoughts Then I guess I may have things backwards in thinking that any personal inconvenience should be outweighed by actual tragety suffered by others. Yeah, not oficially winter, so those people aren't oficially freezing. http://seattletimes.nwsource.com/html/localnews/2003482605_webstormmainbar17.html Or am I also wrong in also thinking that maybe local repair efforts should be aimed at getting people heat, instead of making sure CS players can get online. And no matter what Valve did or didn't do, they are still relying on outside resources as much as anyone else in the area. Even the fact that it may be the Valve employees, or others that are in charge of keeping things running, are too busy keeping their families warm wouldn't be reason enough for some. Maybe we even found out why people feel that CS, DoD, et. al. are loosing players. It may possibly be the lack of soul and community. Last I checked Winter hasn't begun yet. :)~ Obviously people are not complaining about loss of service as being equivalent to loss of life or livelihood... I think you may have gone a bit overboard with that one. The majority of complaints are centered around an apparent lack of re-investment in infrastructure on Valve's part and it is a valid complaint. Perhaps we have demonstrated too much faith in Valves understanding of the most basic concept in network management... that being, no single point failure should bring a network down. That fact stands on it's own and need not be measured against a lost Christmas for those unfortunate to have been effected by the storms. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Sent: Monday, December 18, 2006 12:27 PM To: hlds@list.valvesoftware.com Subject: RE: Re: [hlds] Post-outage thoughts Lets keep things in perspective people. Over a million people without power or heat int the middle of winter. Christmas planning for lots of those are out the window because of lost wages, loss of life, and loss of basic services, and people are whining about not being able to play a computer game for a few days, and then exorcising Valve about not having things back online immediately. Lets think about things based on the effects on real people lives instead of using a tragedy to puff one's resume. It's nice to see the spirit of the holidays on display. Netcom had a massive outage in '96 that lasted almost the same duration as this valve outage. At 11:06 AM 12/18/2006, -Mike- wrote: There are far too many 90% empty datacenters practically sitting on top of major exchanges down here in the SF Bay Area (and all over the US) for Valve to have suffered the outage they saw due to storm conditions. I'm sorry, but a decent distributed network architecture with properly configured load balancing hardware takes care of these single points of failure. But hey, what do I know... I only managed Yahoo's mailservers at GlobalCenter, FriendFinder and Lycos' hardware at Exodus, built and managed bulletproof network backbones at @Home and Netcom... So it's not like I'd know anything about engineering a method of preventing a little lack of power, IP dialtone, or overload from taking your biggest cash machine offline. (sigh) Sorry Valve, I'm gainfully employed and I do not consult on the side. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: Roman Hatsiev [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Sunday, December 17, 2006 11:21:39 PM Subject: Re: [hlds] Post-outage thoughts This is true only as long as you work with tested and widely adopted solution like Active Directory. For closed proprietary system of Steam size designed without redundancy in mind this can be a kind of tricky exercise... Regards, Roman ___ To unsubscribe, edit your list
Re: RE: Re: [hlds] Post-outage thoughts
First off... The finance's mother and extended family lives in West Linn, outside of Portland. They're getting hammered hard. I have some emotional investment in their well being, but you won't see me crying on the list about it. If you get paid well for a job, and you are passionate about your work, you do your job... Even in the most extreme and demanding conditions. There are many electrical linemen and other service workers who are more than aware of this in the Oregon/Washington area at this moment. Anybody who has put in some years in a real commercial telco/network/server operations environment will be quick to tell you the same thing. You do what needs to be done, you keep things running. You Sir, would not be able to handle it. This is the last I have to say on this matter, this is not what this list is for. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: [EMAIL PROTECTED] [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Monday, December 18, 2006 11:42:44 AM Subject: RE: RE: Re: [hlds] Post-outage thoughts Then I guess I may have things backwards in thinking that any personal inconvenience should be outweighed by actual tragety suffered by others. Yeah, not oficially winter, so those people aren't oficially freezing. http://seattletimes.nwsource.com/html/localnews/2003482605_webstormmainbar17.html Or am I also wrong in also thinking that maybe local repair efforts should be aimed at getting people heat, instead of making sure CS players can get online. And no matter what Valve did or didn't do, they are still relying on outside resources as much as anyone else in the area. Even the fact that it may be the Valve employees, or others that are in charge of keeping things running, are too busy keeping their families warm wouldn't be reason enough for some. Maybe we even found out why people feel that CS, DoD, et. al. are loosing players. It may possibly be the lack of soul and community. Last I checked Winter hasn't begun yet. :)~ Obviously people are not complaining about loss of service as being equivalent to loss of life or livelihood... I think you may have gone a bit overboard with that one. The majority of complaints are centered around an apparent lack of re-investment in infrastructure on Valve's part and it is a valid complaint. Perhaps we have demonstrated too much faith in Valves understanding of the most basic concept in network management... that being, no single point failure should bring a network down. That fact stands on it's own and need not be measured against a lost Christmas for those unfortunate to have been effected by the storms. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Sent: Monday, December 18, 2006 12:27 PM To: hlds@list.valvesoftware.com Subject: RE: Re: [hlds] Post-outage thoughts Lets keep things in perspective people. Over a million people without power or heat int the middle of winter. Christmas planning for lots of those are out the window because of lost wages, loss of life, and loss of basic services, and people are whining about not being able to play a computer game for a few days, and then exorcising Valve about not having things back online immediately. Lets think about things based on the effects on real people lives instead of using a tragedy to puff one's resume. It's nice to see the spirit of the holidays on display. Netcom had a massive outage in '96 that lasted almost the same duration as this valve outage. At 11:06 AM 12/18/2006, -Mike- wrote: There are far too many 90% empty datacenters practically sitting on top of major exchanges down here in the SF Bay Area (and all over the US) for Valve to have suffered the outage they saw due to storm conditions. I'm sorry, but a decent distributed network architecture with properly configured load balancing hardware takes care of these single points of failure. But hey, what do I know... I only managed Yahoo's mailservers at GlobalCenter, FriendFinder and Lycos' hardware at Exodus, built and managed bulletproof network backbones at @Home and Netcom... So it's not like I'd know anything about engineering a method of preventing a little lack of power, IP dialtone, or overload from taking your biggest cash machine offline. (sigh) Sorry Valve, I'm gainfully employed and I do not consult on the side. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: Roman Hatsiev [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Sunday, December 17, 2006 11:21:39 PM Subject: Re: [hlds] Post-outage thoughts This is true only as long as you work with tested and widely adopted solution like Active Directory. For closed proprietary system of Steam size designed without redundancy in mind this can be a kind of tricky exercise... Regards, Roman ___ To unsubscribe, edit
RE: RE: RE: RE: Re: [hlds] Post-outage thoughts
I can see from your final post that we are like two ships passing in the night with the one glaring exception that we understand your point of view but I fear you are missing ours entirely. The human interest situation and the network design problems are two distinct and different issues. The common component in both is the storm that caused the power outages and everyone on this list agrees with you that the human tragedy story is more important than the poorly designed network story. If we were discussing the human suffering issue in it's appropriate venue I'm certain you would receive universal agreement with your position. Yes... the suffering is horrible but that has nothing to do with the fact that Valves network was poorly designed. Yes... it is unthinkable that significant numbers of people are without heat with temperatures approaching freezing, but that has nothing to do with the fact that Valve lacked rudimentary network backup. Yes... those that worked hard at restoring the Steam network deserve credit for their hard work but that has nothing to do with the fact that Valve failed to investment spend in an appropriate automated redundancy. You mentioned something about people (on this list) demanding explanations and flashing credentials as if they know step by step and circuit by circuit what happened and why. I don't believe anyone on this list did any such thing. Exactly what happened within the effected area is absolutely irrelevant to the issues being discussed here. There is no dispute that the Valve network failed... that's not the point. The point is that in properly designed network a failure in one location should be detected and compensated for by other locations. This is where Valve failed... not in the storm area, but in Nebraska and Rhode Island and Florida and the UK and Australia and Canada etc. etc. etc. Valve failed to provide for proper redundancy to critical systems and if you were a network specialist you would recognize this inescapable fact immediately and not find fault with those of us who have. We simply must do a better job on this list of understanding what the other guy is trying to tell us. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Sent: Monday, December 18, 2006 3:51 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: RE: Re: [hlds] Post-outage thoughts Ok, last post unless there is a specific point to counter, and a change of tact. I'd like to congratulate everyone at Valve, and especially the people in the IT department for restoring services after a major and widespread weather related disaster. Well done! Such redundancy is Networking 101 and Programming 101... You can choose to ignore it if you like... But in the real word it is fact . Valve is probably making enough money to make it reasonable for them to invest in a redundant system for that money making aparatus. That is Economics 101. You think it looks good to investors that the backbone of the system went down for the entire world because of one geological disaster? You think that's a good selling point for software developers that want to bring their product to market? 273,468 game players couldn't play because Valve had all their eggs in that one geographical basket. Wise business decision? You decide... Ok maybe they are 500 level courses but you still get the point :D -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Monday, December 18, 2006 2:57 PM To: hlds@list.valvesoftware.com Subject: RE: RE: RE: Re: [hlds] Post-outage thoughts All I'm seeing is whining, pettiness, and monday morning quarterbacking. Lets try this. If anyone out there has a diagram of the Valve infrastructure, and a complete understanding of who they contract with for what services and facilities, then lets see it. I only am reading people bitching about what Valve should have done over the last 10 years, and I could do it better, without any reguard or perspective on what the real world impact things may be having in the Seattle area. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: RE: RE: Re: [hlds] Post-outage thoughts
-- [ Picked text/plain from multipart/alternative ] Actually he is right here We were all too stupid to ask whether this eventuality had been catered for by Valve. Well maybe some of you did, but I am embarrassed by the fact that this is a question I've never posed to Valve in the past, as I just ass u me 'ed Valve had dealt with it. Did anybody else bother to query what redundancy Valve had built into the STEAM network? By the looks of things, it is a question that Valve didn't really ask very hard of themselves either, but you live an learn I guess. BTW, it is not just the authentication servers that need to have geographic redundancy, the entire STEAM system and everything that relies on STEAM requires geographic redundancy. eg.VAC, DWP, Purchasing, Friends, hell even the Website and Forums if they want to do the job properly and any other systems that STEAM relies on to function correctly. On 12/19/06, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote: All I'm seeing is whining, pettiness, and monday morning quarterbacking. -- ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: RE: RE: Re: [hlds] Post-outage thoughts
And the hardest part to this is balancing cost versus customer satisfaction. The simple maths of whether it costs more to lose customers due to systems being non-redundant, versus outlaying for data-centre, infrastructure, internet links, etc costs - is what I am assuming they are doing right now. For most of our customers we have BC/ITSC of some sort to cope with this type of situation happening, as it can be very very costly to some organisations, depending on how important IT infrastructure is to them. It's the inevitable insurance policy situation: You are damned if you have it, and damned if you don't... Valve, if you want an Australian presence for your collection of STEAM servers (Colo or fully managed), let me know. I'm sure we can sort something out for you ;) Regards, Adam. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Whisper Sent: Tuesday, 19 December 2006 1:03 PM To: hlds@list.valvesoftware.com Subject: Re: RE: RE: Re: [hlds] Post-outage thoughts -- [ Picked text/plain from multipart/alternative ] Actually he is right here We were all too stupid to ask whether this eventuality had been catered for by Valve. Well maybe some of you did, but I am embarrassed by the fact that this is a question I've never posed to Valve in the past, as I just ass u me 'ed Valve had dealt with it. Did anybody else bother to query what redundancy Valve had built into the STEAM network? By the looks of things, it is a question that Valve didn't really ask very hard of themselves either, but you live an learn I guess. BTW, it is not just the authentication servers that need to have geographic redundancy, the entire STEAM system and everything that relies on STEAM requires geographic redundancy. eg.VAC, DWP, Purchasing, Friends, hell even the Website and Forums if they want to do the job properly and any other systems that STEAM relies on to function correctly. On 12/19/06, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote: All I'm seeing is whining, pettiness, and monday morning quarterbacking. -- ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: [hlds] Post-outage thoughts
-- [ Picked text/plain from multipart/alternative ] Errmmm The problems did not affect all of the banks and ATMs around the US let along around the world, yet the outage stopped ALL steams products. That's the problem. Think global not just regional. From: [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Subject: RE: Re: Re: [hlds] Post-outage thoughts Date: Mon, 18 Dec 2006 14:16:51 -0500 If a power outage that affected over a million poeple in a large geographical area (and is STILL affecting people) isn't a good enough explanation for a short outage, then I'm thinking there may not be one good enough for some around here. I'm reasonably sure that if it was a small yeild nuke, we'd be hearing about how valve servers should have been rad-hardened. Valve was not the only people to go down. Banking, commerce, and other private and public networks also were affected. No company can predict and prepare for every systemic failure that can happen. And yes, I am and administrator. I have been for around 5 years. The only thing that valve OWES me is a REASONABLE attempt to keep things running. And to all the admins out there without hearts three sizes too small. Merry Christmas, and Happy New Year! Honestly, this is an administrators list. Most of us here are going to very coldly and to the fact discuss the impact of the outage. If you are overly sensitive or feel the need for a sleeve to cry on about the human side of the Northwest storm then I'd suggest you unsubscribe and head over to Fark for a group hug. Valve should provide us with a reasonable explanation as to why they have all of their eggs in one geographical basket. They've had nearly ten years to plan and deploy preventative measures against such a failure. -Mike- -Mike- is: Biker ~ Slacker ~ Iconoclast ~ Eclectic Thinker - Original Message From: [EMAIL PROTECTED] [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Sent: Monday, December 18, 2006 9:26:59 AM Subject: RE: Re: [hlds] Post-outage thoughts Lets keep things in perspective people. Over a million people without power or heat int the middle of winter. Christmas planning for lots of those are out the window because of lost wages, loss of life, and loss of basic services, and people are whining about not being able to play a computer game for a few days, and then exorcising Valve about not having things back online immediately. Lets think about things based on the effects on real people lives instead of using a tragedy to puff one's resume. It's nice to see the spirit of the holidays on display. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds _ Be one of the first to try Windows Live Mail. http://ideas.live.com/programpage.aspx?versionId=5d21c51a-b161-4314-9b0e-4911fb2b2e6d -- ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: RE: Re: [hlds] Post-outage thoughts
This absolutely amazing! Next time my mail server goes down for six hours because of lack of UPS in case if I questioned by my management why this happened my reply will be - forget about mail server, just think about the guy who spend these six hours locked in the lift, just think about those spend these six hours without coffee and light! And in case they ask me how these issues are related to the mail server my reply will be - this is your main problem, lack of soul, you bloody lawyers do not care about people and that is why everyone hates you! Hopefully they stop asking stupid question about me not doing my work after that :) Maybe we even found out why people feel that CS, DoD, et. al. are loosing players. It may possibly be the lack of soul and community. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Post-outage thoughts
- Original Message - From: Roman Hatsiev [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Unfortunately things happen - 9/11, New Orleans, you name it - so no single location in a world is good enough for Steam. Yeah but even internic had power after Katrina hit.They where underwater in their parking garage but the websites based out of there still had power,internet ect.It's a large investment called UPS power supply units.For whatever reason the datacenter,I'm thinking more like their home offices, didn't have backup generators.I host most of my servers out of theplanets datacenter and they have multiple backups.Like this listed from their site: DLLSTX2 (3) 500KVA Powerware UPS units, 90 batteries each 1-megawatt generator (2000 gallon tank) 1.5-megawatt generator (2200 gallon tank) DLLSTX4 (2) 500KVA Powerware UPS units, 90 batteries each 1.5-megawatt generator (2200 gallon tank) DLLSTX5 (6) 500KVA Powerware UPS units, 120 batteries each 3.5-megawatts of generating capacity DLLSTX6 (2) 750KVA MGE UPS units (expanding to 6 750 KVA UPS units) 2-megawatt generator (adding two more) And that's just one datacenter company.Of course those are all different buildings in the same town but one company.So if a power outtage hits it may not be able to run for weeks off those units but it would surely be able to power itself until the power company came out.But even the InterNic network was powered for weeks via deliveries of diesel fuel.So as long as you can get gas and keep the generators pumping you've got power. It really boggles my mind that something that is ,in my opinion, a backbone of the steam network would be powered out of one building.I figured they had it spread out from coast to coast and across several continents.But I can imagine they don't trust anyone with the servers that control user auth systems and I can see why.But damn.Have a master that updates the hard files one who bought what ect then simply use slaves that update from the master every 10 minutes or something with just the oks ect.Kinda like a dns server but for gaming auths.Anyways I've gotten long winded. -BeNt- http://www.gorillazsouth.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: [hlds] Post-outage thoughts
Not so long winded... maybe not even long winded enough considering the importance of the topic. I'm an IT director by trade and I can tell you for sure, if what just happened to the Steam network happened to one of the networks I managed... I'd have been out of a job as soon as the network was back up... if not sooner. Even the little two box gamming network I run for my Half-Life servers has UPS backup as does my cable box, routers and switches. I can pop my main breaker and my gamming setup won't miss a beat. I would still like to know why I was unable to play sp Half-Life 2 and sp Half-Life just because the Steam network was down. There shouldn't be any need for authentication of a registered single player game. -Original Message- From: [GS]Admin [mailto:[EMAIL PROTECTED] Sent: Sunday, December 17, 2006 4:54 PM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Post-outage thoughts - Original Message - From: Roman Hatsiev [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Unfortunately things happen - 9/11, New Orleans, you name it - so no single location in a world is good enough for Steam. Yeah but even internic had power after Katrina hit.They where underwater in their parking garage but the websites based out of there still had power,internet ect.It's a large investment called UPS power supply units.For whatever reason the datacenter,I'm thinking more like their home offices, didn't have backup generators.I host most of my servers out of theplanets datacenter and they have multiple backups.Like this listed from their site: DLLSTX2 (3) 500KVA Powerware UPS units, 90 batteries each 1-megawatt generator (2000 gallon tank) 1.5-megawatt generator (2200 gallon tank) DLLSTX4 (2) 500KVA Powerware UPS units, 90 batteries each 1.5-megawatt generator (2200 gallon tank) DLLSTX5 (6) 500KVA Powerware UPS units, 120 batteries each 3.5-megawatts of generating capacity DLLSTX6 (2) 750KVA MGE UPS units (expanding to 6 750 KVA UPS units) 2-megawatt generator (adding two more) And that's just one datacenter company.Of course those are all different buildings in the same town but one company.So if a power outtage hits it may not be able to run for weeks off those units but it would surely be able to power itself until the power company came out.But even the InterNic network was powered for weeks via deliveries of diesel fuel.So as long as you can get gas and keep the generators pumping you've got power. It really boggles my mind that something that is ,in my opinion, a backbone of the steam network would be powered out of one building.I figured they had it spread out from coast to coast and across several continents.But I can imagine they don't trust anyone with the servers that control user auth systems and I can see why.But damn.Have a master that updates the hard files one who bought what ect then simply use slaves that update from the master every 10 minutes or something with just the oks ect.Kinda like a dns server but for gaming auths.Anyways I've gotten long winded. -BeNt- http://www.gorillazsouth.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Post-outage thoughts
-- [ Picked text/plain from multipart/alternative ] An excellent point. There are far too many people who still like to play SP games. This, in my opinion, is totally unacceptable... On 12/17/06, Edward Luna [EMAIL PROTECTED] wrote: I would still like to know why I was unable to play sp Half-Life 2 and sp Half-Life just because the Steam network was down. There shouldn't be any need for authentication of a registered single player game. -- I do what I can. -- ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Post-outage thoughts
-- [ Picked text/plain from multipart/alternative ] Roma, there's one thing you're wrong about - setting up a redundant location with real time replication and automatic failover is not all that complicated and expensive nowadays as one can think. Newbie -Original Message- From: Roman Hatsiev [EMAIL PROTECTED] To: hlds@list.valvesoftware.com Date: Sun, 17 Dec 2006 16:38:18 +0300 Subject: [hlds] Post-outage thoughts First of all I'm sorry for being too pessimistic in the following paragraph, to expect worst must be a general habit of any IT professional. I'm pretty much surprised that the authentication of whole Steam network depends on single location regardless of how good this location is in whatever terms - costs, availability, reliability, etc. Even for my small non-commercial project joining together few thousand members I'm looking for options to introduce some redundancy. Steam on other hand serves millions of customers worldwide and redundancy is a must for the community of such size. Unfortunately things happen - 9/11, New Orleans, you name it - so no single location in a world is good enough for Steam. At the same time I realise that setting up a redundant authentication system is a huge effort in terms of time and money so I suggest introducing premium Steam accounts for some low monthly fee to generate some income for this and other related projects. First obvious advantage to give to premium account is extra bandwidth of content servers and ability to choose a content server to download things from - people generally value the freedom of choice they can get for a reasonable fee :) Second advantage can be a special icon which could be shown is friends chat, score tables and other places where player name is mentioned - kids gonna like it, believe me :) After some brainstorming more advantages which do not require enormous effort can be invented - longer expiration time for some demos, releasing demos/games to premium customers say week before releasing it to everyone else, more guest passes, premium content like wallpapers, desktop themes, etc. By the way, I can generate even more ideas backed with detailed justifications, feel free to ask :) Regards, Roman ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds [http://list.valvesoftware.com/mailman/listinfo/hlds] -- ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
