Re: [AFMUG] bufferbloat
It typically works very well. I only have an issue... maybe 10 times a year. It does get worse when interacting through a CRM that munges everything. - Mike Hammett Intelligent Computing Solutions Midwest Internet Exchange The Brothers WISP - Original Message - From: "Adam Moffett" To: af@af.afmug.com Sent: Monday, February 3, 2020 9:13:01 AM Subject: Re: [AFMUG] bufferbloat That's why I disable threading in my email. It never works anyway. On 2/1/2020 11:26 AM, Mike Hammett wrote: Nothing. Someone forked the conversation into a totally new topic instead of creating a new e-mail. - Mike Hammett Intelligent Computing Solutions Midwest Internet Exchange The Brothers WISP - Original Message - From: "Matt Hoppes" To: "AnimalFarm Microwave Users Group" Sent: Saturday, February 1, 2020 4:28:42 AM Subject: Re: [AFMUG] bufferbloat What does this have to do with CAF-II? For us we pretty much know where the loaded APs are, it’s a matter of smoothing out the experience. On Jan 31, 2020, at 10:18 PM, Simon Westlake < simon@sonar.software > wrote: It's interesting to me how many people are looking at the CoDeL piece of Preseem first - their original vision (and what I still think is the most interesting) is the direct TCP monitoring they do to try to figure out which APs have issues, which customers have issues, and what the root causes of those issues are. The CoDeL piece was just a cherry on top. The CoDeL piece isn't very hard (if you're not worried about scaling it very far) but the monitoring and diagnostic tools they have are, I think, fantastic. I've never heard anyone not rave about them, and I know tons of people using them. On Fri, Jan 31, 2020 at 1:06 PM Ken Hohhof < af...@kwisp.com > wrote: ISP Radio did an interview with them and it’s still on Youtube. 2 years old though. https://www.preseem.com/2018/04/isp-radio-subscriber-queues-latency/ From: AF < af-boun...@af.afmug.com > On Behalf Of David Coudron Sent: Friday, January 31, 2020 12:41 PM To: AnimalFarm Microwave Users Group < af@af.afmug.com > Subject: Re: [AFMUG] bufferbloat Others probably know better. I think some of the other tools are DPI based where Preseem is FQ-CoDel, but I am not sure that is true of Procera. Others integrate to the CRM systems, so I am not sure that is a differentiator. Might be best to hit Preseem up directly on that question. From our experience, they are not high pressure or exaggerative, so you wouldn’t be opening a can of worms you can’t get the lid back on. But maybe someone has done a more in depth investigation of the two. Sorry, wasn’t much help on the question. Regards, David Coudron From: AF < af-boun...@af.afmug.com > On Behalf Of Jason McKemie Sent: Friday, January 31, 2020 11:48 AM To: AnimalFarm Microwave Users Group < af@af.afmug.com > Subject: Re: [AFMUG] bufferbloat How does Preseem compare to Procera? On Fri, Jan 31, 2020 at 11:25 AM Darin Steffl < darin.ste...@mnwifi.com > wrote: I second everything David said. We've been on it 2 years now and it's a tool I will never give up. It's worth every penny On Fri, Jan 31, 2020, 11:03 AM David Coudron < david.coud...@advantenon.com > wrote: We have been using Preseem for about a year now. We originally implemented it as a way to better manage the customer experience and potentially make better use of our DIA bandwidth and maybe reduce some DIA costs. I am guessing someone could build a similar product on their own with open source. However, what we have found is that we get significantly more than the customer experience management with the tool. The reporting is beyond awesome, it has become our number one tool for troubleshooting customers complaints. Others on this list can weigh in on how they use it, but our typical day goes something like this: 1) During our morning Ops call, we take a peek at Preseem's recap of tower latency yesterday. If nothing new shows up for Red towers/access points, we look at Yellow Access Points (this is a ranking of Aps/towers over certain latency thresholds) 2) If any customer calls have come in, we use the Preseem tool to see if they are experiencing latency issues. If they are, we check our SNMP based monitoring tool to see if their wireless connection to the tower has changed or if the AP is experience issues. We had a pretty major windstorm go through two weeks ago, and we found a few customers whose latency spiked and investigation into their connection showed there was an issue with their dish. 3) If latency has climbed, but the AP and upstream devices are all OK, we check into the experience of that customer to others on their tower. Is their latency spike unique, does it happen only under load, etc. More often than not, the issue is specific to them,
Re: [AFMUG] bufferbloat
That's why I disable threading in my email. It never works anyway.
On 2/1/2020 11:26 AM, Mike Hammett wrote:
Nothing. Someone forked the conversation into a totally new topic
instead of creating a new e-mail.
-
Mike Hammett
Intelligent Computing Solutions <http://www.ics-il.com/>
<https://www.facebook.com/ICSIL><https://plus.google.com/+IntelligentComputingSolutionsDeKalb><https://www.linkedin.com/company/intelligent-computing-solutions><https://twitter.com/ICSIL>
Midwest Internet Exchange <http://www.midwest-ix.com/>
<https://www.facebook.com/mdwestix><https://www.linkedin.com/company/midwest-internet-exchange><https://twitter.com/mdwestix>
The Brothers WISP <http://www.thebrotherswisp.com/>
<https://www.facebook.com/thebrotherswisp>
<https://www.youtube.com/channel/UCXSdfxQv7SpoRQYNyLwntZg>
*From: *"Matt Hoppes"
*To: *"AnimalFarm Microwave Users Group"
*Sent: *Saturday, February 1, 2020 4:28:42 AM
*Subject: *Re: [AFMUG] bufferbloat
What does this have to do with CAF-II?
For us we pretty much know where the loaded APs are, it’s a matter of
smoothing out the experience.
On Jan 31, 2020, at 10:18 PM, Simon Westlake <mailto:simon@sonar.software>> wrote:
It's interesting to me how many people are looking at the CoDeL
piece of Preseem first - their original vision (and what I still
think is the most interesting) is the direct TCP monitoring they
do to try to figure out which APs have issues, which customers
have issues, and what the root causes of those issues are. The
CoDeL piece was just a cherry on top.
The CoDeL piece isn't very hard (if you're not worried about
scaling it very far) but the monitoring and diagnostic tools they
have are, I think, fantastic. I've never heard anyone not rave
about them, and I know tons of people using them.
On Fri, Jan 31, 2020 at 1:06 PM Ken Hohhof mailto:af...@kwisp.com>> wrote:
ISP Radio did an interview with them and it’s still on
Youtube. 2 years old though.
https://www.preseem.com/2018/04/isp-radio-subscriber-queues-latency/
*From:* AF mailto:af-boun...@af.afmug.com>> *On Behalf Of *David Coudron
*Sent:* Friday, January 31, 2020 12:41 PM
*To:* AnimalFarm Microwave Users Group mailto:af@af.afmug.com>>
*Subject:* Re: [AFMUG] bufferbloat
Others probably know better. I think some of the other tools
are DPI based where Preseem is FQ-CoDel, but I am not sure
that is true of Procera. Others integrate to the CRM
systems, so I am not sure that is a differentiator. Might be
best to hit Preseem up directly on that question. From our
experience, they are not high pressure or exaggerative, so
you wouldn’t be opening a can of worms you can’t get the lid
back on. But maybe someone has done a more in depth
investigation of the two. Sorry, wasn’t much help on the
question.
Regards,
David Coudron
*From:* AF mailto:af-boun...@af.afmug.com>> *On Behalf Of *Jason McKemie
*Sent:* Friday, January 31, 2020 11:48 AM
*To:* AnimalFarm Microwave Users Group mailto:af@af.afmug.com>>
*Subject:* Re: [AFMUG] bufferbloat
How does Preseem compare to Procera?
On Fri, Jan 31, 2020 at 11:25 AM Darin Steffl
mailto:darin.ste...@mnwifi.com>> wrote:
I second everything David said. We've been on it 2 years
now and it's a tool I will never give up. It's worth every
penny
On Fri, Jan 31, 2020, 11:03 AM David Coudron
mailto:david.coud...@advantenon.com>> wrote:
We have been using Preseem for about a year now. We
originally implemented it as a way to better manage
the customer experience and potentially make better
use of our DIA bandwidth and maybe reduce some DIA
costs. I am guessing someone could build a similar
product on their own with open source.
However, what we have found is that we get
significantly more than the customer experience
management with the tool. The reporting is beyond
awesome, it has become our number one tool for
troubleshooting customers complaints. Others on this
list can weigh in on how they use it, but our typical
day goes something like this:
1) During our morning Ops call, we take a peek at
Preseem's recap of tower latency yesterday. If
nothing new shows up for Red towers/access points, we
look at Yellow Access Points
Re: [AFMUG] bufferbloat
I think there is a community of scolds who criticize what they consider excessively large buffers on the part of ISPs with the pejorative term “bufferbloat”. And yes, there are advanced TCP algorithms today that seek to communicate congestion to the sender and get it to essentially reduce its TCP window size. But I feel like some of the blame should fall on the application vendors, like the streaming video industry. They control the servers and the clients. Why are they advertising such large TCP receive windows in the first place? TCP window scaling broke through the 64 KB ceiling and allows exponential increase to over 1 GB. But why go that far? That’s a lot of data “in the pipe”. Is it fair to criticize “the pipe” (the ISP) for buffering the excess data, even when the receiver says send me more and the sender obliges? The usual argument is to make optimal use of “long fat pipes”, i.e. high bandwidth, high delay connections, to avoid what is called delay-bandwidth product throttling. But what broadband connections would have such high delay, on the order of 500 milliseconds or more, so that leaving that much data unacked would make sense? Only geostationary satellites come to mind. So is this whole “bufferbloat” controversy about streaming video over Hughesnet? I doubt that a 1 second delay in your Super Bowl viewing is the issue. More likely the issue is that other data sharing the same pipe is also delayed, and that data might be latency sensitive – gaming, VoIP, or interactive traffic like web browsing or work at home. And same pipe probably means same customer. So one person in the house watching a video stream from a streaming service or using a streaming app that is aggressive about TCP window size, degrades the Internet experience for others in the same house. The answer may be in algorithms like FQ-CoDel and AQM, and products like Preseem, that not only make the best use of selectively dropping packets to signal congestion to the TCP sender, but use separate queues for each flow and let interactive or latency sensitive traffic bypass whatever mechanism is being used to rate limit bulk flows, whether that mechanism is buffering or packet drop. Because at the end of the day, if you’re going to enforce rate limits, there are only two choices for the nonconforming traffic – buffer, or drop. Or in common terminology shape or police. The alternative is let everything through until the pipes burst. A middle ground is to rate limit the bulk flows as best you can, while giving a pass to small flows that can’t tolerate either delay or packet loss but aren’t big enough to matter in the grand scheme of things. I would not complain if I never heard the term “bufferbloat” again. It targets a legitimate issue, but some people who use the term are just non-thinking scolds. There’s a lot of that these days. If you can pin a nasty name on something, you don’t need to do any heavy thinking about the issue. I know that’s not what anybody in this thread was doing, but the term just kind of sets me off when I hear it. I guess because we are investing a lot of time and money on doing rate limiting in a way that optimizes the user experience, and it’s not just a simple issue, and sometimes I don’t feel the content providers and CDNs are helping very much. From: AF On Behalf Of Mike Hammett Sent: Saturday, February 1, 2020 10:27 AM To: AnimalFarm Microwave Users Group Subject: Re: [AFMUG] bufferbloat Nothing. Someone forked the conversation into a totally new topic instead of creating a new e-mail. - Mike Hammett <http://www.ics-il.com/> Intelligent Computing Solutions <https://www.facebook.com/ICSIL> <https://plus.google.com/+IntelligentComputingSolutionsDeKalb> <https://www.linkedin.com/company/intelligent-computing-solutions> <https://twitter.com/ICSIL> <http://www.midwest-ix.com/> Midwest Internet Exchange <https://www.facebook.com/mdwestix> <https://www.linkedin.com/company/midwest-internet-exchange> <https://twitter.com/mdwestix> <http://www.thebrotherswisp.com/> The Brothers WISP <https://www.facebook.com/thebrotherswisp> <https://www.youtube.com/channel/UCXSdfxQv7SpoRQYNyLwntZg> _ From: "Matt Hoppes" mailto:mattli...@rivervalleyinternet.net> > To: "AnimalFarm Microwave Users Group" mailto:af@af.afmug.com> > Sent: Saturday, February 1, 2020 4:28:42 AM Subject: Re: [AFMUG] bufferbloat What does this have to do with CAF-II? For us we pretty much know where the loaded APs are, it’s a matter of smoothing out the experience. On Jan 31, 2020, at 10:18 PM, Simon Westlake mailto:simon@sonar.software> > wrote: It's interesting to me how many people are looking at the CoDeL piece of Preseem first - their original vision (and what I still think is the most interesting)
Re: [AFMUG] bufferbloat
Nothing. Someone forked the conversation into a totally new topic instead of creating a new e-mail. - Mike Hammett Intelligent Computing Solutions Midwest Internet Exchange The Brothers WISP - Original Message - From: "Matt Hoppes" To: "AnimalFarm Microwave Users Group" Sent: Saturday, February 1, 2020 4:28:42 AM Subject: Re: [AFMUG] bufferbloat What does this have to do with CAF-II? For us we pretty much know where the loaded APs are, it’s a matter of smoothing out the experience. On Jan 31, 2020, at 10:18 PM, Simon Westlake < simon@sonar.software > wrote: It's interesting to me how many people are looking at the CoDeL piece of Preseem first - their original vision (and what I still think is the most interesting) is the direct TCP monitoring they do to try to figure out which APs have issues, which customers have issues, and what the root causes of those issues are. The CoDeL piece was just a cherry on top. The CoDeL piece isn't very hard (if you're not worried about scaling it very far) but the monitoring and diagnostic tools they have are, I think, fantastic. I've never heard anyone not rave about them, and I know tons of people using them. On Fri, Jan 31, 2020 at 1:06 PM Ken Hohhof < af...@kwisp.com > wrote: ISP Radio did an interview with them and it’s still on Youtube. 2 years old though. https://www.preseem.com/2018/04/isp-radio-subscriber-queues-latency/ From: AF < af-boun...@af.afmug.com > On Behalf Of David Coudron Sent: Friday, January 31, 2020 12:41 PM To: AnimalFarm Microwave Users Group < af@af.afmug.com > Subject: Re: [AFMUG] bufferbloat Others probably know better. I think some of the other tools are DPI based where Preseem is FQ-CoDel, but I am not sure that is true of Procera. Others integrate to the CRM systems, so I am not sure that is a differentiator. Might be best to hit Preseem up directly on that question. From our experience, they are not high pressure or exaggerative, so you wouldn’t be opening a can of worms you can’t get the lid back on. But maybe someone has done a more in depth investigation of the two. Sorry, wasn’t much help on the question. Regards, David Coudron From: AF < af-boun...@af.afmug.com > On Behalf Of Jason McKemie Sent: Friday, January 31, 2020 11:48 AM To: AnimalFarm Microwave Users Group < af@af.afmug.com > Subject: Re: [AFMUG] bufferbloat How does Preseem compare to Procera? On Fri, Jan 31, 2020 at 11:25 AM Darin Steffl < darin.ste...@mnwifi.com > wrote: I second everything David said. We've been on it 2 years now and it's a tool I will never give up. It's worth every penny On Fri, Jan 31, 2020, 11:03 AM David Coudron < david.coud...@advantenon.com > wrote: We have been using Preseem for about a year now. We originally implemented it as a way to better manage the customer experience and potentially make better use of our DIA bandwidth and maybe reduce some DIA costs. I am guessing someone could build a similar product on their own with open source. However, what we have found is that we get significantly more than the customer experience management with the tool. The reporting is beyond awesome, it has become our number one tool for troubleshooting customers complaints. Others on this list can weigh in on how they use it, but our typical day goes something like this: 1) During our morning Ops call, we take a peek at Preseem's recap of tower latency yesterday. If nothing new shows up for Red towers/access points, we look at Yellow Access Points (this is a ranking of Aps/towers over certain latency thresholds) 2) If any customer calls have come in, we use the Preseem tool to see if they are experiencing latency issues. If they are, we check our SNMP based monitoring tool to see if their wireless connection to the tower has changed or if the AP is experience issues. We had a pretty major windstorm go through two weeks ago, and we found a few customers whose latency spiked and investigation into their connection showed there was an issue with their dish. 3) If latency has climbed, but the AP and upstream devices are all OK, we check into the experience of that customer to others on their tower. Is their latency spike unique, does it happen only under load, etc. More often than not, the issue is specific to them, doesn't only happen under load, and only at certain times. It is usually from streaming a show on the TV in the far back upstairs bedroom (or something like that) with a crappy connection to their wifi router in the house. We have found it to be an indispensable tool for this kind of thing. We bought it for QoE, but use it daily for monitoring/troubleshooting activities. Not only do you get a hosted reporting solution, you have access to some pretty smart folks. Just this morning our first line of support person said "If Preseem
Re: [AFMUG] bufferbloat
What does this have to do with CAF-II? For us we pretty much know where the loaded APs are, it’s a matter of smoothing out the experience. > On Jan 31, 2020, at 10:18 PM, Simon Westlake wrote: > > It's interesting to me how many people are looking at the CoDeL piece of > Preseem first - their original vision (and what I still think is the most > interesting) is the direct TCP monitoring they do to try to figure out which > APs have issues, which customers have issues, and what the root causes of > those issues are. The CoDeL piece was just a cherry on top. > > The CoDeL piece isn't very hard (if you're not worried about scaling it very > far) but the monitoring and diagnostic tools they have are, I think, > fantastic. I've never heard anyone not rave about them, and I know tons of > people using them. > >> On Fri, Jan 31, 2020 at 1:06 PM Ken Hohhof wrote: >> ISP Radio did an interview with them and it’s still on Youtube. 2 years old >> though. >> >> https://www.preseem.com/2018/04/isp-radio-subscriber-queues-latency/ >> >> >> >> >> >> From: AF On Behalf Of David Coudron >> Sent: Friday, January 31, 2020 12:41 PM >> To: AnimalFarm Microwave Users Group >> Subject: Re: [AFMUG] bufferbloat >> >> >> >> Others probably know better. I think some of the other tools are DPI based >> where Preseem is FQ-CoDel, but I am not sure that is true of Procera. >> Others integrate to the CRM systems, so I am not sure that is a >> differentiator. Might be best to hit Preseem up directly on that question. >> From our experience, they are not high pressure or exaggerative, so you >> wouldn’t be opening a can of worms you can’t get the lid back on. But maybe >> someone has done a more in depth investigation of the two. Sorry, wasn’t >> much help on the question. >> >> >> >> Regards, >> >> >> >> David Coudron >> >> From: AF On Behalf Of Jason McKemie >> Sent: Friday, January 31, 2020 11:48 AM >> To: AnimalFarm Microwave Users Group >> Subject: Re: [AFMUG] bufferbloat >> >> >> >> How does Preseem compare to Procera? >> >> >> >> On Fri, Jan 31, 2020 at 11:25 AM Darin Steffl >> wrote: >> >> I second everything David said. We've been on it 2 years now and it's a tool >> I will never give up. It's worth every penny >> >> >> >> On Fri, Jan 31, 2020, 11:03 AM David Coudron >> wrote: >> >> We have been using Preseem for about a year now. We originally implemented >> it as a way to better manage the customer experience and potentially make >> better use of our DIA bandwidth and maybe reduce some DIA costs. I am >> guessing someone could build a similar product on their own with open >> source. >> >> However, what we have found is that we get significantly more than the >> customer experience management with the tool. The reporting is beyond >> awesome, it has become our number one tool for troubleshooting customers >> complaints. Others on this list can weigh in on how they use it, but our >> typical day goes something like this: >> 1) During our morning Ops call, we take a peek at Preseem's recap of tower >> latency yesterday. If nothing new shows up for Red towers/access points, we >> look at Yellow Access Points (this is a ranking of Aps/towers over certain >> latency thresholds) >> 2) If any customer calls have come in, we use the Preseem tool to see if >> they are experiencing latency issues. If they are, we check our SNMP based >> monitoring tool to see if their wireless connection to the tower has changed >> or if the AP is experience issues. We had a pretty major windstorm go >> through two weeks ago, and we found a few customers whose latency spiked and >> investigation into their connection showed there was an issue with their >> dish. >> 3) If latency has climbed, but the AP and upstream devices are all OK, we >> check into the experience of that customer to others on their tower. Is >> their latency spike unique, does it happen only under load, etc. More >> often than not, the issue is specific to them, doesn't only happen under >> load, and only at certain times. It is usually from streaming a show on the >> TV in the far back upstairs bedroom (or something like that) with a crappy >> connection to their wifi router in the house. >> >> We have found it to be an indispensable tool for this kind of thing. We >> bought it for QoE, but us
Re: [AFMUG] bufferbloat
It's interesting to me how many people are looking at the CoDeL piece of Preseem first - their original vision (and what I still think is the most interesting) is the direct TCP monitoring they do to try to figure out which APs have issues, which customers have issues, and what the root causes of those issues are. The CoDeL piece was just a cherry on top. The CoDeL piece isn't very hard (if you're not worried about scaling it very far) but the monitoring and diagnostic tools they have are, I think, fantastic. I've never heard anyone not rave about them, and I know tons of people using them. On Fri, Jan 31, 2020 at 1:06 PM Ken Hohhof wrote: > ISP Radio did an interview with them and it’s still on Youtube. 2 years > old though. > > https://www.preseem.com/2018/04/isp-radio-subscriber-queues-latency/ > > > > > > *From:* AF *On Behalf Of *David Coudron > *Sent:* Friday, January 31, 2020 12:41 PM > *To:* AnimalFarm Microwave Users Group > *Subject:* Re: [AFMUG] bufferbloat > > > > Others probably know better. I think some of the other tools are DPI > based where Preseem is FQ-CoDel, but I am not sure that is true of > Procera. Others integrate to the CRM systems, so I am not sure that is a > differentiator. Might be best to hit Preseem up directly on that > question. From our experience, they are not high pressure or > exaggerative, so you wouldn’t be opening a can of worms you can’t get the > lid back on. But maybe someone has done a more in depth investigation of > the two. Sorry, wasn’t much help on the question. > > > > Regards, > > > > David Coudron > > *From:* AF *On Behalf Of *Jason McKemie > *Sent:* Friday, January 31, 2020 11:48 AM > *To:* AnimalFarm Microwave Users Group > *Subject:* Re: [AFMUG] bufferbloat > > > > How does Preseem compare to Procera? > > > > On Fri, Jan 31, 2020 at 11:25 AM Darin Steffl > wrote: > > I second everything David said. We've been on it 2 years now and it's a > tool I will never give up. It's worth every penny > > > > On Fri, Jan 31, 2020, 11:03 AM David Coudron > wrote: > > We have been using Preseem for about a year now. We originally > implemented it as a way to better manage the customer experience and > potentially make better use of our DIA bandwidth and maybe reduce some DIA > costs. I am guessing someone could build a similar product on their own > with open source. > > However, what we have found is that we get significantly more than the > customer experience management with the tool. The reporting is beyond > awesome, it has become our number one tool for troubleshooting customers > complaints. Others on this list can weigh in on how they use it, but our > typical day goes something like this: > 1) During our morning Ops call, we take a peek at Preseem's recap of tower > latency yesterday. If nothing new shows up for Red towers/access points, > we look at Yellow Access Points (this is a ranking of Aps/towers over > certain latency thresholds) > 2) If any customer calls have come in, we use the Preseem tool to see if > they are experiencing latency issues. If they are, we check our SNMP > based monitoring tool to see if their wireless connection to the tower has > changed or if the AP is experience issues. We had a pretty major > windstorm go through two weeks ago, and we found a few customers whose > latency spiked and investigation into their connection showed there was an > issue with their dish. > 3) If latency has climbed, but the AP and upstream devices are all OK, we > check into the experience of that customer to others on their tower. Is > their latency spike unique, does it happen only under load, etc. More > often than not, the issue is specific to them, doesn't only happen under > load, and only at certain times. It is usually from streaming a show on > the TV in the far back upstairs bedroom (or something like that) with a > crappy connection to their wifi router in the house. > > We have found it to be an indispensable tool for this kind of thing. We > bought it for QoE, but use it daily for monitoring/troubleshooting > activities. Not only do you get a hosted reporting solution, you have > access to some pretty smart folks. > > Just this morning our first line of support person said "If Preseem ever > goes down, I will cry, it is my favorite troubleshooting tool". We were > having a discussion about how you could compare QoE/Latency from a customer > to other customers on the same AP, to others on the Same Tower, to others > in the same DIA, etc. It is hard to explain how much it changes the way > you think about the "My Internet is slow" complaint. Quite often this > person will get a call ab
Re: [AFMUG] bufferbloat
ISP Radio did an interview with them and it’s still on Youtube. 2 years old though. https://www.preseem.com/2018/04/isp-radio-subscriber-queues-latency/ From: AF On Behalf Of David Coudron Sent: Friday, January 31, 2020 12:41 PM To: AnimalFarm Microwave Users Group Subject: Re: [AFMUG] bufferbloat Others probably know better. I think some of the other tools are DPI based where Preseem is FQ-CoDel, but I am not sure that is true of Procera. Others integrate to the CRM systems, so I am not sure that is a differentiator. Might be best to hit Preseem up directly on that question. From our experience, they are not high pressure or exaggerative, so you wouldn’t be opening a can of worms you can’t get the lid back on. But maybe someone has done a more in depth investigation of the two. Sorry, wasn’t much help on the question. Regards, David Coudron From: AF mailto:af-boun...@af.afmug.com> > On Behalf Of Jason McKemie Sent: Friday, January 31, 2020 11:48 AM To: AnimalFarm Microwave Users Group mailto:af@af.afmug.com> > Subject: Re: [AFMUG] bufferbloat How does Preseem compare to Procera? On Fri, Jan 31, 2020 at 11:25 AM Darin Steffl mailto:darin.ste...@mnwifi.com> > wrote: I second everything David said. We've been on it 2 years now and it's a tool I will never give up. It's worth every penny On Fri, Jan 31, 2020, 11:03 AM David Coudron mailto:david.coud...@advantenon.com> > wrote: We have been using Preseem for about a year now. We originally implemented it as a way to better manage the customer experience and potentially make better use of our DIA bandwidth and maybe reduce some DIA costs. I am guessing someone could build a similar product on their own with open source. However, what we have found is that we get significantly more than the customer experience management with the tool. The reporting is beyond awesome, it has become our number one tool for troubleshooting customers complaints. Others on this list can weigh in on how they use it, but our typical day goes something like this: 1) During our morning Ops call, we take a peek at Preseem's recap of tower latency yesterday. If nothing new shows up for Red towers/access points, we look at Yellow Access Points (this is a ranking of Aps/towers over certain latency thresholds) 2) If any customer calls have come in, we use the Preseem tool to see if they are experiencing latency issues. If they are, we check our SNMP based monitoring tool to see if their wireless connection to the tower has changed or if the AP is experience issues. We had a pretty major windstorm go through two weeks ago, and we found a few customers whose latency spiked and investigation into their connection showed there was an issue with their dish. 3) If latency has climbed, but the AP and upstream devices are all OK, we check into the experience of that customer to others on their tower. Is their latency spike unique, does it happen only under load, etc. More often than not, the issue is specific to them, doesn't only happen under load, and only at certain times. It is usually from streaming a show on the TV in the far back upstairs bedroom (or something like that) with a crappy connection to their wifi router in the house. We have found it to be an indispensable tool for this kind of thing. We bought it for QoE, but use it daily for monitoring/troubleshooting activities. Not only do you get a hosted reporting solution, you have access to some pretty smart folks. Just this morning our first line of support person said "If Preseem ever goes down, I will cry, it is my favorite troubleshooting tool". We were having a discussion about how you could compare QoE/Latency from a customer to other customers on the same AP, to others on the Same Tower, to others in the same DIA, etc. It is hard to explain how much it changes the way you think about the "My Internet is slow" complaint. Quite often this person will get a call about it being slow last night, and she will ask the time at which it happened and pull up very detailed information like "You were using 45 of you 50 Mbps plan with 50 ms latency". Take the time to go through the demo with Gerrit. You may not decide it is not for you, but it won't be a waste of time to understand why they are pushing it so hard. David Coudron david.coud...@advantenon.com <mailto:david.coud...@advantenon.com> | Mobile: 612-991-7474 Advantenon, Inc. i...@advantenon.com <mailto:i...@advantenon.com> | 3500 Vicksburg Lane N, Suite 315, Plymouth, MN 55447 | www.advantenon.com <http://www.advantenon.com> | Phone: 800-704-4720 | Local: 612-454-1545 -Original Message- From: AF mailto:af-boun...@af.afmug.com> > On Behalf Of Adam Moffett Sent: Friday, January 31, 2020 10:41 AM To: af@af.afmug.com <mailto:
Re: [AFMUG] bufferbloat
Others probably know better. I think some of the other tools are DPI based where Preseem is FQ-CoDel, but I am not sure that is true of Procera. Others integrate to the CRM systems, so I am not sure that is a differentiator. Might be best to hit Preseem up directly on that question. From our experience, they are not high pressure or exaggerative, so you wouldn’t be opening a can of worms you can’t get the lid back on. But maybe someone has done a more in depth investigation of the two. Sorry, wasn’t much help on the question. Regards, David Coudron From: AF On Behalf Of Jason McKemie Sent: Friday, January 31, 2020 11:48 AM To: AnimalFarm Microwave Users Group Subject: Re: [AFMUG] bufferbloat How does Preseem compare to Procera? On Fri, Jan 31, 2020 at 11:25 AM Darin Steffl mailto:darin.ste...@mnwifi.com>> wrote: I second everything David said. We've been on it 2 years now and it's a tool I will never give up. It's worth every penny On Fri, Jan 31, 2020, 11:03 AM David Coudron mailto:david.coud...@advantenon.com>> wrote: We have been using Preseem for about a year now. We originally implemented it as a way to better manage the customer experience and potentially make better use of our DIA bandwidth and maybe reduce some DIA costs. I am guessing someone could build a similar product on their own with open source. However, what we have found is that we get significantly more than the customer experience management with the tool. The reporting is beyond awesome, it has become our number one tool for troubleshooting customers complaints. Others on this list can weigh in on how they use it, but our typical day goes something like this: 1) During our morning Ops call, we take a peek at Preseem's recap of tower latency yesterday. If nothing new shows up for Red towers/access points, we look at Yellow Access Points (this is a ranking of Aps/towers over certain latency thresholds) 2) If any customer calls have come in, we use the Preseem tool to see if they are experiencing latency issues. If they are, we check our SNMP based monitoring tool to see if their wireless connection to the tower has changed or if the AP is experience issues. We had a pretty major windstorm go through two weeks ago, and we found a few customers whose latency spiked and investigation into their connection showed there was an issue with their dish. 3) If latency has climbed, but the AP and upstream devices are all OK, we check into the experience of that customer to others on their tower. Is their latency spike unique, does it happen only under load, etc. More often than not, the issue is specific to them, doesn't only happen under load, and only at certain times. It is usually from streaming a show on the TV in the far back upstairs bedroom (or something like that) with a crappy connection to their wifi router in the house. We have found it to be an indispensable tool for this kind of thing. We bought it for QoE, but use it daily for monitoring/troubleshooting activities. Not only do you get a hosted reporting solution, you have access to some pretty smart folks. Just this morning our first line of support person said "If Preseem ever goes down, I will cry, it is my favorite troubleshooting tool". We were having a discussion about how you could compare QoE/Latency from a customer to other customers on the same AP, to others on the Same Tower, to others in the same DIA, etc. It is hard to explain how much it changes the way you think about the "My Internet is slow" complaint. Quite often this person will get a call about it being slow last night, and she will ask the time at which it happened and pull up very detailed information like "You were using 45 of you 50 Mbps plan with 50 ms latency". Take the time to go through the demo with Gerrit. You may not decide it is not for you, but it won't be a waste of time to understand why they are pushing it so hard. David Coudron david.coud...@advantenon.com<mailto:david.coud...@advantenon.com> | Mobile: 612-991-7474 Advantenon, Inc. i...@advantenon.com<mailto:i...@advantenon.com> | 3500 Vicksburg Lane N, Suite 315, Plymouth, MN 55447 | www.advantenon.com<http://www.advantenon.com> | Phone: 800-704-4720 | Local: 612-454-1545 -Original Message- From: AF mailto:af-boun...@af.afmug.com>> On Behalf Of Adam Moffett Sent: Friday, January 31, 2020 10:41 AM To: af@af.afmug.com<mailto:af@af.afmug.com> Subject: Re: [AFMUG] bufferbloat I think they have integration with common CRM's like Sonar. You sound exactly like I sounded 15 years ago. The more stuff I have to deal with every day, the more I'm ok with outsourcing some of my troubles to someone else. I just paid a guy $800 to replace an exhaust inducer in my furnace. I know that inducer is $99 and goes in with 4 screws and a hose clamp, but it's more wort
Re: [AFMUG] bufferbloat
You know, maybe I was thinking of Procera not Preseem with regards to TCP monitoring On Fri, Jan 31, 2020 at 9:57 AM Ken Hohhof wrote: > Can you imagine doing coding back in these days: > > https://arstechnica.com/science/2020/01/a-deep-dive-into-the-apollo-guidance-computer-and-the-hack-that-saved-apollo-14/ > > Talk about both time and hardware constraints, oh and people die if you > screw up. I wonder what their effective cost per hour (or line of code) > was. > > > -Original Message- > From: AF On Behalf Of Robert > Sent: Friday, January 31, 2020 11:39 AM > To: af@af.afmug.com > Subject: Re: [AFMUG] bufferbloat > > When I was a wee lad working at my first real engineering job in nuclear > power, my boss had me sit down and figure out what my actual self-cost was > as my very first job for him. I came up with a number like $22/hour. He > then showed me what my actual overhead number was, and it came out to like > $58 for the company and he told me that the company would be billing me out > at over $100. I was astonished. So it's an exercise I still do. Now I > am at about $95 all up cost. Much of the work I do is at a loss against > that, but I make it up in bulk, LOL. But the time value of money is more > like $450 right now. The hours left in life are getting shorter, bank > assets are getting longer. Spend more time on fun than work. > > On 1/31/20 8:41 AM, Adam Moffett wrote: > > I think they have integration with common CRM's like Sonar. > > > > You sound exactly like I sounded 15 years ago. The more stuff I have > > to deal with every day, the more I'm ok with outsourcing some of my > > troubles to someone else. > > > > I just paid a guy $800 to replace an exhaust inducer in my furnace. I > > know that inducer is $99 and goes in with 4 screws and a hose clamp, > > but it's more worth my time to let someone else take care of it so I > > can do something else. Same goes for Preseem vs the $300 Linux box. > > > > I'm not knocking your method. There's a point in the business cycle > > where there's more time than there is cash, and it will make sense to > > do some more DIY things. I'm just saying the Preseem thing has value > > too. > > > > -Adam > > > > > > On 1/31/2020 11:34 AM, Dev wrote: > >> I’m getting spammed like every day with the Preseem guys selling what > >> seem like expensive hacks of fq_codel to reduce bufferbloat. Is there > >> anything else interesting about their technology besides deploying > >> open source implementation of fq_codel or CAKE on commodity hardware, > >> which we already do to great effect on a $300 single board Linux box > >> with a few ports? I guess they have a pretty dashboard, anyhing other > >> than that? > > > > > -- > AF mailing list > AF@af.afmug.com > http://af.afmug.com/mailman/listinfo/af_af.afmug.com > > > > -- > AF mailing list > AF@af.afmug.com > http://af.afmug.com/mailman/listinfo/af_af.afmug.com > -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com
Re: [AFMUG] bufferbloat
Can you imagine doing coding back in these days: https://arstechnica.com/science/2020/01/a-deep-dive-into-the-apollo-guidance-computer-and-the-hack-that-saved-apollo-14/ Talk about both time and hardware constraints, oh and people die if you screw up. I wonder what their effective cost per hour (or line of code) was. -Original Message- From: AF On Behalf Of Robert Sent: Friday, January 31, 2020 11:39 AM To: af@af.afmug.com Subject: Re: [AFMUG] bufferbloat When I was a wee lad working at my first real engineering job in nuclear power, my boss had me sit down and figure out what my actual self-cost was as my very first job for him. I came up with a number like $22/hour. He then showed me what my actual overhead number was, and it came out to like $58 for the company and he told me that the company would be billing me out at over $100. I was astonished. So it's an exercise I still do. Now I am at about $95 all up cost. Much of the work I do is at a loss against that, but I make it up in bulk, LOL. But the time value of money is more like $450 right now. The hours left in life are getting shorter, bank assets are getting longer. Spend more time on fun than work. On 1/31/20 8:41 AM, Adam Moffett wrote: > I think they have integration with common CRM's like Sonar. > > You sound exactly like I sounded 15 years ago. The more stuff I have > to deal with every day, the more I'm ok with outsourcing some of my > troubles to someone else. > > I just paid a guy $800 to replace an exhaust inducer in my furnace. I > know that inducer is $99 and goes in with 4 screws and a hose clamp, > but it's more worth my time to let someone else take care of it so I > can do something else. Same goes for Preseem vs the $300 Linux box. > > I'm not knocking your method. There's a point in the business cycle > where there's more time than there is cash, and it will make sense to > do some more DIY things. I'm just saying the Preseem thing has value > too. > > -Adam > > > On 1/31/2020 11:34 AM, Dev wrote: >> I’m getting spammed like every day with the Preseem guys selling what >> seem like expensive hacks of fq_codel to reduce bufferbloat. Is there >> anything else interesting about their technology besides deploying >> open source implementation of fq_codel or CAKE on commodity hardware, >> which we already do to great effect on a $300 single board Linux box >> with a few ports? I guess they have a pretty dashboard, anyhing other >> than that? > -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com
Re: [AFMUG] bufferbloat
Or I guess more appropriately now, Sandvine / Procera. On Fri, Jan 31, 2020 at 11:48 AM Jason McKemie < j.mcke...@veloxinetbroadband.com> wrote: > How does Preseem compare to Procera? > > On Fri, Jan 31, 2020 at 11:25 AM Darin Steffl > wrote: > >> I second everything David said. We've been on it 2 years now and it's a >> tool I will never give up. It's worth every penny >> >> On Fri, Jan 31, 2020, 11:03 AM David Coudron < >> david.coud...@advantenon.com> wrote: >> >>> We have been using Preseem for about a year now. We originally >>> implemented it as a way to better manage the customer experience and >>> potentially make better use of our DIA bandwidth and maybe reduce some DIA >>> costs. I am guessing someone could build a similar product on their own >>> with open source. >>> >>> However, what we have found is that we get significantly more than the >>> customer experience management with the tool. The reporting is beyond >>> awesome, it has become our number one tool for troubleshooting customers >>> complaints. Others on this list can weigh in on how they use it, but our >>> typical day goes something like this: >>> 1) During our morning Ops call, we take a peek at Preseem's recap of >>> tower latency yesterday. If nothing new shows up for Red towers/access >>> points, we look at Yellow Access Points (this is a ranking of Aps/towers >>> over certain latency thresholds) >>> 2) If any customer calls have come in, we use the Preseem tool to see if >>> they are experiencing latency issues. If they are, we check our SNMP >>> based monitoring tool to see if their wireless connection to the tower has >>> changed or if the AP is experience issues. We had a pretty major >>> windstorm go through two weeks ago, and we found a few customers whose >>> latency spiked and investigation into their connection showed there was an >>> issue with their dish. >>> 3) If latency has climbed, but the AP and upstream devices are all OK, >>> we check into the experience of that customer to others on their tower. >>> Is their latency spike unique, does it happen only under load, etc. More >>> often than not, the issue is specific to them, doesn't only happen under >>> load, and only at certain times. It is usually from streaming a show on >>> the TV in the far back upstairs bedroom (or something like that) with a >>> crappy connection to their wifi router in the house. >>> >>> We have found it to be an indispensable tool for this kind of thing. >>> We bought it for QoE, but use it daily for monitoring/troubleshooting >>> activities. Not only do you get a hosted reporting solution, you have >>> access to some pretty smart folks. >>> >>> Just this morning our first line of support person said "If Preseem ever >>> goes down, I will cry, it is my favorite troubleshooting tool". We were >>> having a discussion about how you could compare QoE/Latency from a customer >>> to other customers on the same AP, to others on the Same Tower, to others >>> in the same DIA, etc. It is hard to explain how much it changes the way >>> you think about the "My Internet is slow" complaint. Quite often this >>> person will get a call about it being slow last night, and she will ask the >>> time at which it happened and pull up very detailed information like "You >>> were using 45 of you 50 Mbps plan with 50 ms latency". >>> >>> Take the time to go through the demo with Gerrit. You may not decide >>> it is not for you, but it won't be a waste of time to understand why they >>> are pushing it so hard. >>> >>> David Coudron >>> david.coud...@advantenon.com | Mobile: 612-991-7474 >>> >>> Advantenon, Inc. >>> i...@advantenon.com | 3500 Vicksburg Lane N, Suite 315, Plymouth, MN >>> 55447 | www.advantenon.com | Phone: 800-704-4720 | Local: >>> 612-454-1545 >>> >>> >>> >>> -Original Message- >>> From: AF On Behalf Of Adam Moffett >>> Sent: Friday, January 31, 2020 10:41 AM >>> To: af@af.afmug.com >>> Subject: Re: [AFMUG] bufferbloat >>> >>> I think they have integration with common CRM's like Sonar. >>> >>> You sound exactly like I sounded 15 years ago. The more stuff I have to >>> deal with every day, the more I'm ok with outsourcing some of my troubles >>> to someone else. >
Re: [AFMUG] bufferbloat
How does Preseem compare to Procera? On Fri, Jan 31, 2020 at 11:25 AM Darin Steffl wrote: > I second everything David said. We've been on it 2 years now and it's a > tool I will never give up. It's worth every penny > > On Fri, Jan 31, 2020, 11:03 AM David Coudron > wrote: > >> We have been using Preseem for about a year now. We originally >> implemented it as a way to better manage the customer experience and >> potentially make better use of our DIA bandwidth and maybe reduce some DIA >> costs. I am guessing someone could build a similar product on their own >> with open source. >> >> However, what we have found is that we get significantly more than the >> customer experience management with the tool. The reporting is beyond >> awesome, it has become our number one tool for troubleshooting customers >> complaints. Others on this list can weigh in on how they use it, but our >> typical day goes something like this: >> 1) During our morning Ops call, we take a peek at Preseem's recap of >> tower latency yesterday. If nothing new shows up for Red towers/access >> points, we look at Yellow Access Points (this is a ranking of Aps/towers >> over certain latency thresholds) >> 2) If any customer calls have come in, we use the Preseem tool to see if >> they are experiencing latency issues. If they are, we check our SNMP >> based monitoring tool to see if their wireless connection to the tower has >> changed or if the AP is experience issues. We had a pretty major >> windstorm go through two weeks ago, and we found a few customers whose >> latency spiked and investigation into their connection showed there was an >> issue with their dish. >> 3) If latency has climbed, but the AP and upstream devices are all OK, we >> check into the experience of that customer to others on their tower. Is >> their latency spike unique, does it happen only under load, etc. More >> often than not, the issue is specific to them, doesn't only happen under >> load, and only at certain times. It is usually from streaming a show on >> the TV in the far back upstairs bedroom (or something like that) with a >> crappy connection to their wifi router in the house. >> >> We have found it to be an indispensable tool for this kind of thing. We >> bought it for QoE, but use it daily for monitoring/troubleshooting >> activities. Not only do you get a hosted reporting solution, you have >> access to some pretty smart folks. >> >> Just this morning our first line of support person said "If Preseem ever >> goes down, I will cry, it is my favorite troubleshooting tool". We were >> having a discussion about how you could compare QoE/Latency from a customer >> to other customers on the same AP, to others on the Same Tower, to others >> in the same DIA, etc. It is hard to explain how much it changes the way >> you think about the "My Internet is slow" complaint. Quite often this >> person will get a call about it being slow last night, and she will ask the >> time at which it happened and pull up very detailed information like "You >> were using 45 of you 50 Mbps plan with 50 ms latency". >> >> Take the time to go through the demo with Gerrit. You may not decide it >> is not for you, but it won't be a waste of time to understand why they are >> pushing it so hard. >> >> David Coudron >> david.coud...@advantenon.com | Mobile: 612-991-7474 >> >> Advantenon, Inc. >> i...@advantenon.com | 3500 Vicksburg Lane N, Suite 315, Plymouth, MN >> 55447 | www.advantenon.com | Phone: 800-704-4720 | Local: >> 612-454-1545 >> >> >> >> -Original Message- >> From: AF On Behalf Of Adam Moffett >> Sent: Friday, January 31, 2020 10:41 AM >> To: af@af.afmug.com >> Subject: Re: [AFMUG] bufferbloat >> >> I think they have integration with common CRM's like Sonar. >> >> You sound exactly like I sounded 15 years ago. The more stuff I have to >> deal with every day, the more I'm ok with outsourcing some of my troubles >> to someone else. >> >> I just paid a guy $800 to replace an exhaust inducer in my furnace. I >> know that inducer is $99 and goes in with 4 screws and a hose clamp, but >> it's more worth my time to let someone else take care of it so I can do >> something else. Same goes for Preseem vs the $300 Linux box. >> >> I'm not knocking your method. There's a point in the business cycle >> where there's more time than there is cash, and it will make sense to do >> some more DIY things. I'm just sa
Re: [AFMUG] bufferbloat
I’m not sure I want a ton more WISPs to sign up with them. Their support is very good, and if they get a bunch more customers, they will either have to hire a bunch more people or start outsourcing like everybody else. On the other hand, they’re not getting rich on just what they charge me, so maybe I do need more people to sign up so they stay in business. From: AF On Behalf Of David Coudron Sent: Friday, January 31, 2020 11:34 AM To: AnimalFarm Microwave Users Group Subject: Re: [AFMUG] bufferbloat There are others on here that can explain it far better than me. However, here’s a shot: 1. The Preseem system doesn’t use Deep Packet Inspection. It uses FQ-CoDel to manage the queues to ensure better experience for the customer. So they can prioritize traffic that needs instant response like clicking a link on a web page versus filling the buffer on the Netflix client. 2. Customers are soft limited to their plan amounts. By this I mean that their traffic is managed more as they approach their plan limits which reduces the abrupt web page not responding types if issues if their streaming is gobbling up their connection. 3. Customer speed plans are managed in the Preseem appliance rather than Mikrotik queues. This is significantly simpler when integrating with Sonar, Powercode and others. Very much a plug and play installation I know there is more to this technically, but I wouldn’t do a good job of diving into it as we didn’t need/want to get that deep into it. We just needed to confirm it did what they claimed. Again, we bought it for the traffic shaping, but the pleasant surprise was the troubleshooting tools with it. I think our surprise wasn’t that we got some extra tools, but how accurate and useful they were. It is hard to oversell that. Regards, David Coudron From: AF mailto:af-boun...@af.afmug.com> > On Behalf Of Dan Spitler Sent: Friday, January 31, 2020 11:20 AM To: AnimalFarm Microwave Users Group mailto:af@af.afmug.com> > Subject: Re: [AFMUG] bufferbloat I'm more interested in the QoE monitoring. I'm guessing it looks at TCP performance? How does it do it? Deep packet inspection? Seems tough to do without overwhelming the server or increasing latency significantly. On Fri, Jan 31, 2020 at 9:03 AM David Coudron mailto:david.coud...@advantenon.com> > wrote: We have been using Preseem for about a year now. We originally implemented it as a way to better manage the customer experience and potentially make better use of our DIA bandwidth and maybe reduce some DIA costs. I am guessing someone could build a similar product on their own with open source. However, what we have found is that we get significantly more than the customer experience management with the tool. The reporting is beyond awesome, it has become our number one tool for troubleshooting customers complaints. Others on this list can weigh in on how they use it, but our typical day goes something like this: 1) During our morning Ops call, we take a peek at Preseem's recap of tower latency yesterday. If nothing new shows up for Red towers/access points, we look at Yellow Access Points (this is a ranking of Aps/towers over certain latency thresholds) 2) If any customer calls have come in, we use the Preseem tool to see if they are experiencing latency issues. If they are, we check our SNMP based monitoring tool to see if their wireless connection to the tower has changed or if the AP is experience issues. We had a pretty major windstorm go through two weeks ago, and we found a few customers whose latency spiked and investigation into their connection showed there was an issue with their dish. 3) If latency has climbed, but the AP and upstream devices are all OK, we check into the experience of that customer to others on their tower. Is their latency spike unique, does it happen only under load, etc. More often than not, the issue is specific to them, doesn't only happen under load, and only at certain times. It is usually from streaming a show on the TV in the far back upstairs bedroom (or something like that) with a crappy connection to their wifi router in the house. We have found it to be an indispensable tool for this kind of thing. We bought it for QoE, but use it daily for monitoring/troubleshooting activities. Not only do you get a hosted reporting solution, you have access to some pretty smart folks. Just this morning our first line of support person said "If Preseem ever goes down, I will cry, it is my favorite troubleshooting tool". We were having a discussion about how you could compare QoE/Latency from a customer to other customers on the same AP, to others on the Same Tower, to others in the same DIA, etc. It is hard to explain how much it changes the way you think about the "My Internet is slow" complaint. Quite of
Re: [AFMUG] bufferbloat
When I was a wee lad working at my first real engineering job in nuclear power, my boss had me sit down and figure out what my actual self-cost was as my very first job for him. I came up with a number like $22/hour. He then showed me what my actual overhead number was, and it came out to like $58 for the company and he told me that the company would be billing me out at over $100. I was astonished. So it's an exercise I still do. Now I am at about $95 all up cost. Much of the work I do is at a loss against that, but I make it up in bulk, LOL. But the time value of money is more like $450 right now. The hours left in life are getting shorter, bank assets are getting longer. Spend more time on fun than work. On 1/31/20 8:41 AM, Adam Moffett wrote: I think they have integration with common CRM's like Sonar. You sound exactly like I sounded 15 years ago. The more stuff I have to deal with every day, the more I'm ok with outsourcing some of my troubles to someone else. I just paid a guy $800 to replace an exhaust inducer in my furnace. I know that inducer is $99 and goes in with 4 screws and a hose clamp, but it's more worth my time to let someone else take care of it so I can do something else. Same goes for Preseem vs the $300 Linux box. I'm not knocking your method. There's a point in the business cycle where there's more time than there is cash, and it will make sense to do some more DIY things. I'm just saying the Preseem thing has value too. -Adam On 1/31/2020 11:34 AM, Dev wrote: I’m getting spammed like every day with the Preseem guys selling what seem like expensive hacks of fq_codel to reduce bufferbloat. Is there anything else interesting about their technology besides deploying open source implementation of fq_codel or CAKE on commodity hardware, which we already do to great effect on a $300 single board Linux box with a few ports? I guess they have a pretty dashboard, anyhing other than that? -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com
Re: [AFMUG] bufferbloat
No DPI. Uses flow behavior and active queue management. If it acts interactive, it is treated as such. If it acts like a bulk flow, it is treated as such. I assume it’s smart enough that if Netflix opens up 4 parallel TCP connections from different ports numbers or different IP addresses, it’s treated as one flow, but I don’t know the nuts and bolts well enough to say. But it’s not going to give you analytics or have you create policies for Netflix vs Disney+ vs Windows Update. Or streaming vs gaming vs VoIP. That is implicit not explicit. From: AF On Behalf Of Dan Spitler Sent: Friday, January 31, 2020 11:20 AM To: AnimalFarm Microwave Users Group Subject: Re: [AFMUG] bufferbloat I'm more interested in the QoE monitoring. I'm guessing it looks at TCP performance? How does it do it? Deep packet inspection? Seems tough to do without overwhelming the server or increasing latency significantly. On Fri, Jan 31, 2020 at 9:03 AM David Coudron mailto:david.coud...@advantenon.com> > wrote: We have been using Preseem for about a year now. We originally implemented it as a way to better manage the customer experience and potentially make better use of our DIA bandwidth and maybe reduce some DIA costs. I am guessing someone could build a similar product on their own with open source. However, what we have found is that we get significantly more than the customer experience management with the tool. The reporting is beyond awesome, it has become our number one tool for troubleshooting customers complaints. Others on this list can weigh in on how they use it, but our typical day goes something like this: 1) During our morning Ops call, we take a peek at Preseem's recap of tower latency yesterday. If nothing new shows up for Red towers/access points, we look at Yellow Access Points (this is a ranking of Aps/towers over certain latency thresholds) 2) If any customer calls have come in, we use the Preseem tool to see if they are experiencing latency issues. If they are, we check our SNMP based monitoring tool to see if their wireless connection to the tower has changed or if the AP is experience issues. We had a pretty major windstorm go through two weeks ago, and we found a few customers whose latency spiked and investigation into their connection showed there was an issue with their dish. 3) If latency has climbed, but the AP and upstream devices are all OK, we check into the experience of that customer to others on their tower. Is their latency spike unique, does it happen only under load, etc. More often than not, the issue is specific to them, doesn't only happen under load, and only at certain times. It is usually from streaming a show on the TV in the far back upstairs bedroom (or something like that) with a crappy connection to their wifi router in the house. We have found it to be an indispensable tool for this kind of thing. We bought it for QoE, but use it daily for monitoring/troubleshooting activities. Not only do you get a hosted reporting solution, you have access to some pretty smart folks. Just this morning our first line of support person said "If Preseem ever goes down, I will cry, it is my favorite troubleshooting tool". We were having a discussion about how you could compare QoE/Latency from a customer to other customers on the same AP, to others on the Same Tower, to others in the same DIA, etc. It is hard to explain how much it changes the way you think about the "My Internet is slow" complaint. Quite often this person will get a call about it being slow last night, and she will ask the time at which it happened and pull up very detailed information like "You were using 45 of you 50 Mbps plan with 50 ms latency". Take the time to go through the demo with Gerrit. You may not decide it is not for you, but it won't be a waste of time to understand why they are pushing it so hard. David Coudron david.coud...@advantenon.com <mailto:david.coud...@advantenon.com> | Mobile: 612-991-7474 Advantenon, Inc. i...@advantenon.com <mailto:i...@advantenon.com> | 3500 Vicksburg Lane N, Suite 315, Plymouth, MN 55447 | www.advantenon.com <http://www.advantenon.com> | Phone: 800-704-4720 | Local: 612-454-1545 -Original Message- From: AF mailto:af-boun...@af.afmug.com> > On Behalf Of Adam Moffett Sent: Friday, January 31, 2020 10:41 AM To: af@af.afmug.com <mailto:af@af.afmug.com> Subject: Re: [AFMUG] bufferbloat I think they have integration with common CRM's like Sonar. You sound exactly like I sounded 15 years ago. The more stuff I have to deal with every day, the more I'm ok with outsourcing some of my troubles to someone else. I just paid a guy $800 to replace an exhaust inducer in my furnace. I know that inducer is $99 and goes in with 4 screws and a hose clamp, but it's more worth
Re: [AFMUG] bufferbloat
There are others on here that can explain it far better than me. However, here’s a shot: 1. The Preseem system doesn’t use Deep Packet Inspection. It uses FQ-CoDel to manage the queues to ensure better experience for the customer.So they can prioritize traffic that needs instant response like clicking a link on a web page versus filling the buffer on the Netflix client. 2. Customers are soft limited to their plan amounts. By this I mean that their traffic is managed more as they approach their plan limits which reduces the abrupt web page not responding types if issues if their streaming is gobbling up their connection. 3. Customer speed plans are managed in the Preseem appliance rather than Mikrotik queues. This is significantly simpler when integrating with Sonar, Powercode and others. Very much a plug and play installation I know there is more to this technically, but I wouldn’t do a good job of diving into it as we didn’t need/want to get that deep into it. We just needed to confirm it did what they claimed. Again, we bought it for the traffic shaping, but the pleasant surprise was the troubleshooting tools with it. I think our surprise wasn’t that we got some extra tools, but how accurate and useful they were. It is hard to oversell that. Regards, David Coudron From: AF On Behalf Of Dan Spitler Sent: Friday, January 31, 2020 11:20 AM To: AnimalFarm Microwave Users Group Subject: Re: [AFMUG] bufferbloat I'm more interested in the QoE monitoring. I'm guessing it looks at TCP performance? How does it do it? Deep packet inspection? Seems tough to do without overwhelming the server or increasing latency significantly. On Fri, Jan 31, 2020 at 9:03 AM David Coudron mailto:david.coud...@advantenon.com>> wrote: We have been using Preseem for about a year now. We originally implemented it as a way to better manage the customer experience and potentially make better use of our DIA bandwidth and maybe reduce some DIA costs. I am guessing someone could build a similar product on their own with open source. However, what we have found is that we get significantly more than the customer experience management with the tool. The reporting is beyond awesome, it has become our number one tool for troubleshooting customers complaints. Others on this list can weigh in on how they use it, but our typical day goes something like this: 1) During our morning Ops call, we take a peek at Preseem's recap of tower latency yesterday. If nothing new shows up for Red towers/access points, we look at Yellow Access Points (this is a ranking of Aps/towers over certain latency thresholds) 2) If any customer calls have come in, we use the Preseem tool to see if they are experiencing latency issues. If they are, we check our SNMP based monitoring tool to see if their wireless connection to the tower has changed or if the AP is experience issues. We had a pretty major windstorm go through two weeks ago, and we found a few customers whose latency spiked and investigation into their connection showed there was an issue with their dish. 3) If latency has climbed, but the AP and upstream devices are all OK, we check into the experience of that customer to others on their tower. Is their latency spike unique, does it happen only under load, etc. More often than not, the issue is specific to them, doesn't only happen under load, and only at certain times. It is usually from streaming a show on the TV in the far back upstairs bedroom (or something like that) with a crappy connection to their wifi router in the house. We have found it to be an indispensable tool for this kind of thing. We bought it for QoE, but use it daily for monitoring/troubleshooting activities. Not only do you get a hosted reporting solution, you have access to some pretty smart folks. Just this morning our first line of support person said "If Preseem ever goes down, I will cry, it is my favorite troubleshooting tool". We were having a discussion about how you could compare QoE/Latency from a customer to other customers on the same AP, to others on the Same Tower, to others in the same DIA, etc. It is hard to explain how much it changes the way you think about the "My Internet is slow" complaint. Quite often this person will get a call about it being slow last night, and she will ask the time at which it happened and pull up very detailed information like "You were using 45 of you 50 Mbps plan with 50 ms latency". Take the time to go through the demo with Gerrit. You may not decide it is not for you, but it won't be a waste of time to understand why they are pushing it so hard. David Coudron david.coud...@advantenon.com<mailto:david.coud...@advantenon.com> | Mobile: 612-991-7474 Advantenon, Inc. i...@advantenon.com<mailto:i...@advantenon.com> | 3500 Vicksburg Lane N, Suite 315, Plymouth, MN 55447
Re: [AFMUG] bufferbloat
Mike brings up a good point about integration. It found all our Cambium APs and pulled all sorts of information from them and figured out which customers were on which APs, without us doing anything. We use PPPoE and dynamic pools and they use the Mikrotik API to map PPPoE sessions to customer IP addresses. Very minimal work on our part to give the Preseem appliance access to the Mikrotik API, and it just worked. And we don’t even use Sonar, if we did, they have integration with Sonar and it could pull stuff like customer speed plans and DHCP assignments. From: AF On Behalf Of Mike Hammett Sent: Friday, January 31, 2020 10:40 AM To: AnimalFarm Microwave Users Group Subject: Re: [AFMUG] bufferbloat Please send a fresh e-mail when starting a new thread. Anybody can do nearly anything most of the manufacturers do with a $300 PC, open source software, and sufficient time. A) Is it worth your time instead of just paying them? B) Are you qualified to maintain the implementation? C) Is the user experience as nice? D) Have you built all of the integrations they have? - Mike Hammett <http://www.ics-il.com/> Intelligent Computing Solutions <https://www.facebook.com/ICSIL> <https://plus.google.com/+IntelligentComputingSolutionsDeKalb> <https://www.linkedin.com/company/intelligent-computing-solutions> <https://twitter.com/ICSIL> <http://www.midwest-ix.com/> Midwest Internet Exchange <https://www.facebook.com/mdwestix> <https://www.linkedin.com/company/midwest-internet-exchange> <https://twitter.com/mdwestix> <http://www.thebrotherswisp.com/> The Brothers WISP <https://www.facebook.com/thebrotherswisp> <https://www.youtube.com/channel/UCXSdfxQv7SpoRQYNyLwntZg> _ From: "Dev" mailto:d...@logicalwebhost.com> > To: "AnimalFarm Microwave Users Group" mailto:af@af.afmug.com> > Sent: Friday, January 31, 2020 10:34:39 AM Subject: [AFMUG] bufferbloat I’m getting spammed like every day with the Preseem guys selling what seem like expensive hacks of fq_codel to reduce bufferbloat. Is there anything else interesting about their technology besides deploying open source implementation of fq_codel or CAKE on commodity hardware, which we already do to great effect on a $300 single board Linux box with a few ports? I guess they have a pretty dashboard, anyhing other than that? -- AF mailing list AF@af.afmug.com <mailto:AF@af.afmug.com> http://af.afmug.com/mailman/listinfo/af_af.afmug.com -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com
Re: [AFMUG] bufferbloat
I second everything David said. We've been on it 2 years now and it's a tool I will never give up. It's worth every penny On Fri, Jan 31, 2020, 11:03 AM David Coudron wrote: > We have been using Preseem for about a year now. We originally > implemented it as a way to better manage the customer experience and > potentially make better use of our DIA bandwidth and maybe reduce some DIA > costs. I am guessing someone could build a similar product on their own > with open source. > > However, what we have found is that we get significantly more than the > customer experience management with the tool. The reporting is beyond > awesome, it has become our number one tool for troubleshooting customers > complaints. Others on this list can weigh in on how they use it, but our > typical day goes something like this: > 1) During our morning Ops call, we take a peek at Preseem's recap of tower > latency yesterday. If nothing new shows up for Red towers/access points, > we look at Yellow Access Points (this is a ranking of Aps/towers over > certain latency thresholds) > 2) If any customer calls have come in, we use the Preseem tool to see if > they are experiencing latency issues. If they are, we check our SNMP > based monitoring tool to see if their wireless connection to the tower has > changed or if the AP is experience issues. We had a pretty major > windstorm go through two weeks ago, and we found a few customers whose > latency spiked and investigation into their connection showed there was an > issue with their dish. > 3) If latency has climbed, but the AP and upstream devices are all OK, we > check into the experience of that customer to others on their tower. Is > their latency spike unique, does it happen only under load, etc. More > often than not, the issue is specific to them, doesn't only happen under > load, and only at certain times. It is usually from streaming a show on > the TV in the far back upstairs bedroom (or something like that) with a > crappy connection to their wifi router in the house. > > We have found it to be an indispensable tool for this kind of thing. We > bought it for QoE, but use it daily for monitoring/troubleshooting > activities. Not only do you get a hosted reporting solution, you have > access to some pretty smart folks. > > Just this morning our first line of support person said "If Preseem ever > goes down, I will cry, it is my favorite troubleshooting tool". We were > having a discussion about how you could compare QoE/Latency from a customer > to other customers on the same AP, to others on the Same Tower, to others > in the same DIA, etc. It is hard to explain how much it changes the way > you think about the "My Internet is slow" complaint. Quite often this > person will get a call about it being slow last night, and she will ask the > time at which it happened and pull up very detailed information like "You > were using 45 of you 50 Mbps plan with 50 ms latency". > > Take the time to go through the demo with Gerrit. You may not decide it > is not for you, but it won't be a waste of time to understand why they are > pushing it so hard. > > David Coudron > david.coud...@advantenon.com | Mobile: 612-991-7474 > > Advantenon, Inc. > i...@advantenon.com | 3500 Vicksburg Lane N, Suite 315, Plymouth, MN > 55447 | www.advantenon.com | Phone: 800-704-4720 | Local: > 612-454-1545 > > > > -Original Message- > From: AF On Behalf Of Adam Moffett > Sent: Friday, January 31, 2020 10:41 AM > To: af@af.afmug.com > Subject: Re: [AFMUG] bufferbloat > > I think they have integration with common CRM's like Sonar. > > You sound exactly like I sounded 15 years ago. The more stuff I have to > deal with every day, the more I'm ok with outsourcing some of my troubles > to someone else. > > I just paid a guy $800 to replace an exhaust inducer in my furnace. I > know that inducer is $99 and goes in with 4 screws and a hose clamp, but > it's more worth my time to let someone else take care of it so I can do > something else. Same goes for Preseem vs the $300 Linux box. > > I'm not knocking your method. There's a point in the business cycle where > there's more time than there is cash, and it will make sense to do some > more DIY things. I'm just saying the Preseem thing has value too. > > -Adam > > > On 1/31/2020 11:34 AM, Dev wrote: > > I’m getting spammed like every day with the Preseem guys selling what > seem like expensive hacks of fq_codel to reduce bufferbloat. Is there > anything else interesting about their technology besides deploying open > source implementation of fq_codel or CAKE on commodity hardware, which we > already do to great
Re: [AFMUG] bufferbloat
I'm more interested in the QoE monitoring. I'm guessing it looks at TCP performance? How does it do it? Deep packet inspection? Seems tough to do without overwhelming the server or increasing latency significantly. On Fri, Jan 31, 2020 at 9:03 AM David Coudron wrote: > We have been using Preseem for about a year now. We originally > implemented it as a way to better manage the customer experience and > potentially make better use of our DIA bandwidth and maybe reduce some DIA > costs. I am guessing someone could build a similar product on their own > with open source. > > However, what we have found is that we get significantly more than the > customer experience management with the tool. The reporting is beyond > awesome, it has become our number one tool for troubleshooting customers > complaints. Others on this list can weigh in on how they use it, but our > typical day goes something like this: > 1) During our morning Ops call, we take a peek at Preseem's recap of tower > latency yesterday. If nothing new shows up for Red towers/access points, > we look at Yellow Access Points (this is a ranking of Aps/towers over > certain latency thresholds) > 2) If any customer calls have come in, we use the Preseem tool to see if > they are experiencing latency issues. If they are, we check our SNMP > based monitoring tool to see if their wireless connection to the tower has > changed or if the AP is experience issues. We had a pretty major > windstorm go through two weeks ago, and we found a few customers whose > latency spiked and investigation into their connection showed there was an > issue with their dish. > 3) If latency has climbed, but the AP and upstream devices are all OK, we > check into the experience of that customer to others on their tower. Is > their latency spike unique, does it happen only under load, etc. More > often than not, the issue is specific to them, doesn't only happen under > load, and only at certain times. It is usually from streaming a show on > the TV in the far back upstairs bedroom (or something like that) with a > crappy connection to their wifi router in the house. > > We have found it to be an indispensable tool for this kind of thing. We > bought it for QoE, but use it daily for monitoring/troubleshooting > activities. Not only do you get a hosted reporting solution, you have > access to some pretty smart folks. > > Just this morning our first line of support person said "If Preseem ever > goes down, I will cry, it is my favorite troubleshooting tool". We were > having a discussion about how you could compare QoE/Latency from a customer > to other customers on the same AP, to others on the Same Tower, to others > in the same DIA, etc. It is hard to explain how much it changes the way > you think about the "My Internet is slow" complaint. Quite often this > person will get a call about it being slow last night, and she will ask the > time at which it happened and pull up very detailed information like "You > were using 45 of you 50 Mbps plan with 50 ms latency". > > Take the time to go through the demo with Gerrit. You may not decide it > is not for you, but it won't be a waste of time to understand why they are > pushing it so hard. > > David Coudron > david.coud...@advantenon.com | Mobile: 612-991-7474 > > Advantenon, Inc. > i...@advantenon.com | 3500 Vicksburg Lane N, Suite 315, Plymouth, MN > 55447 | www.advantenon.com | Phone: 800-704-4720 | Local: > 612-454-1545 > > > > -Original Message- > From: AF On Behalf Of Adam Moffett > Sent: Friday, January 31, 2020 10:41 AM > To: af@af.afmug.com > Subject: Re: [AFMUG] bufferbloat > > I think they have integration with common CRM's like Sonar. > > You sound exactly like I sounded 15 years ago. The more stuff I have to > deal with every day, the more I'm ok with outsourcing some of my troubles > to someone else. > > I just paid a guy $800 to replace an exhaust inducer in my furnace. I > know that inducer is $99 and goes in with 4 screws and a hose clamp, but > it's more worth my time to let someone else take care of it so I can do > something else. Same goes for Preseem vs the $300 Linux box. > > I'm not knocking your method. There's a point in the business cycle where > there's more time than there is cash, and it will make sense to do some > more DIY things. I'm just saying the Preseem thing has value too. > > -Adam > > > On 1/31/2020 11:34 AM, Dev wrote: > > I’m getting spammed like every day with the Preseem guys selling what > seem like expensive hacks of fq_codel to reduce bufferbloat. Is there > anything else interesting about their technology besides deploying open > sou
Re: [AFMUG] bufferbloat
We have been using Preseem for about a year now. We originally implemented it as a way to better manage the customer experience and potentially make better use of our DIA bandwidth and maybe reduce some DIA costs. I am guessing someone could build a similar product on their own with open source. However, what we have found is that we get significantly more than the customer experience management with the tool. The reporting is beyond awesome, it has become our number one tool for troubleshooting customers complaints. Others on this list can weigh in on how they use it, but our typical day goes something like this: 1) During our morning Ops call, we take a peek at Preseem's recap of tower latency yesterday. If nothing new shows up for Red towers/access points, we look at Yellow Access Points (this is a ranking of Aps/towers over certain latency thresholds) 2) If any customer calls have come in, we use the Preseem tool to see if they are experiencing latency issues. If they are, we check our SNMP based monitoring tool to see if their wireless connection to the tower has changed or if the AP is experience issues. We had a pretty major windstorm go through two weeks ago, and we found a few customers whose latency spiked and investigation into their connection showed there was an issue with their dish. 3) If latency has climbed, but the AP and upstream devices are all OK, we check into the experience of that customer to others on their tower. Is their latency spike unique, does it happen only under load, etc. More often than not, the issue is specific to them, doesn't only happen under load, and only at certain times. It is usually from streaming a show on the TV in the far back upstairs bedroom (or something like that) with a crappy connection to their wifi router in the house. We have found it to be an indispensable tool for this kind of thing. We bought it for QoE, but use it daily for monitoring/troubleshooting activities. Not only do you get a hosted reporting solution, you have access to some pretty smart folks. Just this morning our first line of support person said "If Preseem ever goes down, I will cry, it is my favorite troubleshooting tool". We were having a discussion about how you could compare QoE/Latency from a customer to other customers on the same AP, to others on the Same Tower, to others in the same DIA, etc. It is hard to explain how much it changes the way you think about the "My Internet is slow" complaint. Quite often this person will get a call about it being slow last night, and she will ask the time at which it happened and pull up very detailed information like "You were using 45 of you 50 Mbps plan with 50 ms latency". Take the time to go through the demo with Gerrit. You may not decide it is not for you, but it won't be a waste of time to understand why they are pushing it so hard. David Coudron david.coud...@advantenon.com | Mobile: 612-991-7474 Advantenon, Inc. i...@advantenon.com | 3500 Vicksburg Lane N, Suite 315, Plymouth, MN 55447 | www.advantenon.com | Phone: 800-704-4720 | Local: 612-454-1545 -Original Message- From: AF On Behalf Of Adam Moffett Sent: Friday, January 31, 2020 10:41 AM To: af@af.afmug.com Subject: Re: [AFMUG] bufferbloat I think they have integration with common CRM's like Sonar. You sound exactly like I sounded 15 years ago. The more stuff I have to deal with every day, the more I'm ok with outsourcing some of my troubles to someone else. I just paid a guy $800 to replace an exhaust inducer in my furnace. I know that inducer is $99 and goes in with 4 screws and a hose clamp, but it's more worth my time to let someone else take care of it so I can do something else. Same goes for Preseem vs the $300 Linux box. I'm not knocking your method. There's a point in the business cycle where there's more time than there is cash, and it will make sense to do some more DIY things. I'm just saying the Preseem thing has value too. -Adam On 1/31/2020 11:34 AM, Dev wrote: > I’m getting spammed like every day with the Preseem guys selling what seem > like expensive hacks of fq_codel to reduce bufferbloat. Is there anything > else interesting about their technology besides deploying open source > implementation of fq_codel or CAKE on commodity hardware, which we already do > to great effect on a $300 single board Linux box with a few ports? I guess > they have a pretty dashboard, anyhing other than that? -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com
Re: [AFMUG] bufferbloat
Keep hoping that Mikrotik v7 will have codel. If you look at there forum many requests for it. -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com
Re: [AFMUG] bufferbloat
On 1/31/20 8:49 AM, Ken Hohhof wrote: Well, yes, but you will probably dismiss it as stuff you can implement yourself plus a "pretty dashboard". Most stuff these days is based on Linux and cloud management, and it's a question of whether you want to pay someone to do it, or spend your own time on development and maintenance. You could dismiss Sonar or cnMaestro as just a pretty dashboard, but I don't think that's fair. I've become old enough that I appreciate the time savings that buying a thing can offer more now than when I was younger. There was a time when I'd stand up a custom server and install linux just to have a file server, now I'll just buy a Synology so I can spend time on other things since. Also back when I was first building linux file servers there weren't any linux-based NAS appliances, but now there are so that's a task I can skip if I choose to. -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com
Re: [AFMUG] bufferbloat
I think they have integration with common CRM's like Sonar. You sound exactly like I sounded 15 years ago. The more stuff I have to deal with every day, the more I'm ok with outsourcing some of my troubles to someone else. I just paid a guy $800 to replace an exhaust inducer in my furnace. I know that inducer is $99 and goes in with 4 screws and a hose clamp, but it's more worth my time to let someone else take care of it so I can do something else. Same goes for Preseem vs the $300 Linux box. I'm not knocking your method. There's a point in the business cycle where there's more time than there is cash, and it will make sense to do some more DIY things. I'm just saying the Preseem thing has value too. -Adam On 1/31/2020 11:34 AM, Dev wrote: I’m getting spammed like every day with the Preseem guys selling what seem like expensive hacks of fq_codel to reduce bufferbloat. Is there anything else interesting about their technology besides deploying open source implementation of fq_codel or CAKE on commodity hardware, which we already do to great effect on a $300 single board Linux box with a few ports? I guess they have a pretty dashboard, anyhing other than that? -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com
Re: [AFMUG] bufferbloat
Please send a fresh e-mail when starting a new thread. Anybody can do nearly anything most of the manufacturers do with a $300 PC, open source software, and sufficient time. A) Is it worth your time instead of just paying them? B) Are you qualified to maintain the implementation? C) Is the user experience as nice? D) Have you built all of the integrations they have? - Mike Hammett Intelligent Computing Solutions Midwest Internet Exchange The Brothers WISP - Original Message - From: "Dev" To: "AnimalFarm Microwave Users Group" Sent: Friday, January 31, 2020 10:34:39 AM Subject: [AFMUG] bufferbloat I’m getting spammed like every day with the Preseem guys selling what seem like expensive hacks of fq_codel to reduce bufferbloat. Is there anything else interesting about their technology besides deploying open source implementation of fq_codel or CAKE on commodity hardware, which we already do to great effect on a $300 single board Linux box with a few ports? I guess they have a pretty dashboard, anyhing other than that? -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com
[AFMUG] bufferbloat
I’m getting spammed like every day with the Preseem guys selling what seem like expensive hacks of fq_codel to reduce bufferbloat. Is there anything else interesting about their technology besides deploying open source implementation of fq_codel or CAKE on commodity hardware, which we already do to great effect on a $300 single board Linux box with a few ports? I guess they have a pretty dashboard, anyhing other than that? -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com
