Re: [Gluster-users] 40 gig ethernet
On 21 Jun 2013, at 14:00, Shawn Nock wrote: And why do makers of RAID cards make it so hard to update firmware? They persist in requiring DOS, Java or even Windows, I almost always have to resort to some unsupported hack in order to get updates done on Linux. I'm pretty sure with the 3ware controllers(or at least most of the newer ones from the 9xxx series) you can flash under Linux with some CLI utility. If I remember correctly one of our 3ware SAS controllers even had an update button on the 3dm2 webpanel. Marcus ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
On 21 Jun 2013, at 14:00, Shawn Nock wrote: > I had to keep a stock of spares in-house until I migrated to 3ware (now > LSI). I haven't had any trouble with these cards in several years (and > haven't needed to RMA or contact support). I've got a 3Ware 9650SE-8LPML SATA RAID controller that's been a bit troublesome. It was working fine but died on a scheduled reboot, in such a way that even the BIOS wouldn't POST! 3Ware were good about replacing it, but the replacement they sent was DOA, the second one worked ok. I still find reboots on this machine very stressful! And why do makers of RAID cards make it so hard to update firmware? They persist in requiring DOS, Java or even Windows, I almost always have to resort to some unsupported hack in order to get updates done on Linux. Marcus ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
Weird, I have a bunch of servers with Areca ARC-1680 8-ports and they have never given me a problem. The first thing I did was update the firmware to the latest - my brand new cards had firmware 2 years old - and didn't recognize disks > 1TB. On Thu, Jun 20, 2013 at 7:11 AM, Shawn Nock wrote: > Justin Clift writes: >>> The other issue I have is with hardware RAID. I'm not sure if folks >>> are using that with gluster or if they're using software RAID, but >>> the closed source nature and crappy proprietary tools annoys all the >>> devops guys I know. What are you all doing for your gluster setups? >>> Is there some magical RAID controller that has Free tools, or are >>> people using mdadm, or are people just unhappy or ? >> >> Before joining Red Hat I was using Areca hardware. But Areca (the >> company) was weird/dishonest when I tried to RMA a card that went bad. >> >> So, I advise people to keep away from that crowd. Haven't tried any >> others in depth since. :/ > > I second the thoughts on Areca. They are a terrible company; avoid at > all costs. I've RMA'd every card I've installed of theirs that had been > in service for more that 6 months, some servers have had RMA returns > fail within months. > > Their only US support option is "we'll ship it to Taiwan for repair and > return it is 6-8 weeks". There is no option to pay for advanced > replacement. > > I had to keep a stock of spares in-house until I migrated to 3ware (now > LSI). I haven't had any trouble with these cards in several years (and > haven't needed to RMA or contact support). > > -- > Shawn Nock (OpenPGP: 0x65118FA5) > > ___ > Gluster-users mailing list > Gluster-users@gluster.org > http://supercolony.gluster.org/mailman/listinfo/gluster-users ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
Justin Clift writes: >> The other issue I have is with hardware RAID. I'm not sure if folks >> are using that with gluster or if they're using software RAID, but >> the closed source nature and crappy proprietary tools annoys all the >> devops guys I know. What are you all doing for your gluster setups? >> Is there some magical RAID controller that has Free tools, or are >> people using mdadm, or are people just unhappy or ? > > Before joining Red Hat I was using Areca hardware. But Areca (the > company) was weird/dishonest when I tried to RMA a card that went bad. > > So, I advise people to keep away from that crowd. Haven't tried any > others in depth since. :/ I second the thoughts on Areca. They are a terrible company; avoid at all costs. I've RMA'd every card I've installed of theirs that had been in service for more that 6 months, some servers have had RMA returns fail within months. Their only US support option is "we'll ship it to Taiwan for repair and return it is 6-8 weeks". There is no option to pay for advanced replacement. I had to keep a stock of spares in-house until I migrated to 3ware (now LSI). I haven't had any trouble with these cards in several years (and haven't needed to RMA or contact support). -- Shawn Nock (OpenPGP: 0x65118FA5) pgpyp9RwsWROw.pgp Description: PGP signature ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
On Wed, Jun 19, 2013 at 4:00 PM, Justin Clift wrote: > > > >> PS: FWIW I wrote a puppet module to manage LSI RAID. It drove me crazy >> using their tool on some supermicro hardware I had. If anyone shows >> interest, I can post the code. > > That corresponds to this blog post doesn't it? :) > > https://ttboj.wordpress.com/2013/06/17/puppet-lsi-hardware-raid-module/ Yup, just posted it after some people emailed me asking for it. Hope it helps. Feedback welcome. James ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
On 17/06/2013, at 4:01 AM, James wrote: > I have to jump in here and add that I'm with you for the drivers > aspect. I had a lot of problems with the 10gE drivers when getting > gluster going. I haven't tested recently, but it's a huge worry when > buying hardware. Even RedHat had a lot of trouble confirming if > certain chips would work! That's good to know about. In my personal home dev/test lab here I'm just using Mellanox DDR ConnectX cards ($39 on eBay! :>) and running things with either IPoIB or in RDMA mode. I did try switching the cards into 10GbE mode (worked fine), but don't really see the point of running these cards at half speed and worse (10GbE) in a home lab. :) > The other issue I have is with hardware RAID. I'm not sure if folks > are using that with gluster or if they're using software RAID, but the > closed source nature and crappy proprietary tools annoys all the > devops guys I know. What are you all doing for your gluster setups? Is > there some magical RAID controller that has Free tools, or are people > using mdadm, or are people just unhappy or ? Before joining Red Hat I was using Areca hardware. But Areca (the company) was weird/dishonest when I tried to RMA a card that went bad. So, I advise people to keep away from that crowd. Haven't tried any others in depth since. :/ > PS: FWIW I wrote a puppet module to manage LSI RAID. It drove me crazy > using their tool on some supermicro hardware I had. If anyone shows > interest, I can post the code. That corresponds to this blog post doesn't it? :) https://ttboj.wordpress.com/2013/06/17/puppet-lsi-hardware-raid-module/ Regards and best wishes, Justin Clift -- Open Source and Standards @ Red Hat twitter.com/realjustinclift ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
I'm using the inbuilt Infiniband drivers that come with CentOS 6.x. I did go through the pain of downloading an ISO from Mellanox and installing all their specially built tools, went through their tuning guide, and saw no speed improvements at all. the IPoIB module cannot push the speeds like native RDMA - but I've not been able to get gluster to work with infiniband correctly. (Get massive CPU spikes from glusterd, filesystem stalls, and terrible speeds - bassically native rdma was unusable). I've not tried the 3.4 branch yet (my native rdma attempts have all been with the 3.3.x series). Anyway, I can completely blow out the raw speed of my underlying RAID10 arrays across my boxes with IPoIB/Infiniband so it doesn't matter. I chose Infiniband because overall it is far cheaper than 10G cards and associated switches (2 years ago). Prices have no moved enough for me to bother with 10G. On Sat, Jun 15, 2013 at 5:34 PM, Justin Clift wrote: > On 14/06/2013, at 8:13 PM, Bryan Whitehead wrote: >> I'm using 40G Infiniband with IPoIB for gluster. Here are some ping >> times (from host 172.16.1.10): >> >> [root@node0.cloud ~]# ping -c 10 172.16.1.11 >> PING 172.16.1.11 (172.16.1.11) 56(84) bytes of data. >> 64 bytes from 172.16.1.11: icmp_seq=1 ttl=64 time=0.093 ms >> 64 bytes from 172.16.1.11: icmp_seq=2 ttl=64 time=0.113 ms >> 64 bytes from 172.16.1.11: icmp_seq=3 ttl=64 time=0.163 ms >> 64 bytes from 172.16.1.11: icmp_seq=4 ttl=64 time=0.125 ms >> 64 bytes from 172.16.1.11: icmp_seq=5 ttl=64 time=0.125 ms >> 64 bytes from 172.16.1.11: icmp_seq=6 ttl=64 time=0.125 ms >> 64 bytes from 172.16.1.11: icmp_seq=7 ttl=64 time=0.198 ms >> 64 bytes from 172.16.1.11: icmp_seq=8 ttl=64 time=0.171 ms >> 64 bytes from 172.16.1.11: icmp_seq=9 ttl=64 time=0.194 ms >> 64 bytes from 172.16.1.11: icmp_seq=10 ttl=64 time=0.115 ms > > > Out of curiosity, are you using connected mode or datagram mode > for this? Also, are you using the inbuilt OS infiniband drivers, > or Mellanox's OFED? (Or Intel/QLogic's equivalent if using > their stuff) > > Asking because I haven't yet seen any real "best practise" stuff > on ways to set this up for Gluster (yet). ;) > > Regards and best wishes, > > Justin Clift > > -- > Open Source and Standards @ Red Hat > > twitter.com/realjustinclift > ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
On Sun, Jun 16, 2013 at 9:03 PM, sal poliandro wrote: > On a side not the native linux drivers have not really kept up with the 40gb > cards. Linux still has issues with some 10gb cards. If you are going 40gb > talk to the people that license the dna driver. They are in paramus nj and > do a lot of higer end networks with proper Linux drivers. The media (dac > cables or om4 mtp) dont seem to affect performance much as long as you dont > push the dac longer than 3-5 meeters. I have to jump in here and add that I'm with you for the drivers aspect. I had a lot of problems with the 10gE drivers when getting gluster going. I haven't tested recently, but it's a huge worry when buying hardware. Even RedHat had a lot of trouble confirming if certain chips would work! The other issue I have is with hardware RAID. I'm not sure if folks are using that with gluster or if they're using software RAID, but the closed source nature and crappy proprietary tools annoys all the devops guys I know. What are you all doing for your gluster setups? Is there some magical RAID controller that has Free tools, or are people using mdadm, or are people just unhappy or ? Cheers, James PS: FWIW I wrote a puppet module to manage LSI RAID. It drove me crazy using their tool on some supermicro hardware I had. If anyone shows interest, I can post the code. ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
Most of the 40gb stuff is designed for mostly East/West traffic as that tends to be the majority of traffic in the datacenter these days. All the big guys make platforms that can keep full port to port across the platform between 4-7. 40gb has not fallen that far where it is not still a decent size investment to do right and as someone who keeps trying to fit gluster into production I have found that other storage platforms always beat out gluster in top end hardware. When 40gb is more high end and the 100gb starts to take marketshare gluster may work in some enviornments but when running top of the line network and server gear the TCO of a commerical storage product (and the support that comes with it) always wins, at least for me. On a side not the native linux drivers have not really kept up with the 40gb cards. Linux still has issues with some 10gb cards. If you are going 40gb talk to the people that license the dna driver. They are in paramus nj and do a lot of higer end networks with proper Linux drivers. The media (dac cables or om4 mtp) dont seem to affect performance much as long as you dont push the dac longer than 3-5 meeters. Salvatore "Popsikle" Poliandro Sent from my mobile, please excuse any typos. One day we will have mobile devices where we don't need this footer :) On Jun 14, 2013 10:04 AM, "Nathan Stratton" wrote: > I have been playing around with Gluster on and off for the last 6 years or > so. Most of the things that have been keeping me from using it have been > related to latency. > > In the past I have been using 10 gig infiniband or 10 gig ethernet, > recently the price of 40 gig ethernet has fallen quite a bit with guys like > Arista. > > My question is, is this worth it at all for something like Gluster? The > port to port latency looks impressive at under 4 microseconds, but I don't > yet know what total system to system latency would look like assuming QSPF+ > copper cables and linux stack. > > -- > ><> > Nathan Stratton Founder, CTO > Exario Networks, Inc. > nathan at robotics.net nathan at > exarionetworks.com > http://www.robotics.net > http://www.exarionetworks.com/ > > Building the WebRTC solutions today that your customers will demand > tomorrow. > > ___ > Gluster-users mailing list > Gluster-users@gluster.org > http://supercolony.gluster.org/mailman/listinfo/gluster-users > ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
On Fri, Jun 14, 2013 at 2:13 PM, Bryan Whitehead wrote: > I'm using 40G Infiniband with IPoIB for gluster. Here are some ping > times (from host 172.16.1.10): > > --- 172.16.1.11 ping statistics --- > 10 packets transmitted, 10 received, 0% packet loss, time 8999ms > rtt min/avg/max/mdev = 0.093/0.142/0.198/0.035 ms Interesting, you have a lower min and slightly lower avg, but your max is actually higher then I am seeing on my 10 gig setup, since Gluster FS uses a lot of small packets, it does not look like it is worth upgrading from 10 to 40 gig ethernet... --- virt1.exarionetworks.com ping statistics --- 10 packets transmitted, 10 received, 0% packet loss, time 8999ms rtt min/avg/max/mdev = 0.112/0.163/0.191/0.027 ms Also note, that I am running 10GBase-T, if I was running over fiber, I would think my overall numbers would be lower then your 40 gig Infiniband, how can that be? -- ><> Nathan Stratton Founder, CTO Exario Networks, Inc. nathan at robotics.net nathan at exarionetworks.com http://www.robotics.net http://www.exarionetworks.com/ Building the WebRTC solutions today that your customers will demand tomorrow. ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
On 14/06/2013, at 8:13 PM, Bryan Whitehead wrote: > I'm using 40G Infiniband with IPoIB for gluster. Here are some ping > times (from host 172.16.1.10): > > [root@node0.cloud ~]# ping -c 10 172.16.1.11 > PING 172.16.1.11 (172.16.1.11) 56(84) bytes of data. > 64 bytes from 172.16.1.11: icmp_seq=1 ttl=64 time=0.093 ms > 64 bytes from 172.16.1.11: icmp_seq=2 ttl=64 time=0.113 ms > 64 bytes from 172.16.1.11: icmp_seq=3 ttl=64 time=0.163 ms > 64 bytes from 172.16.1.11: icmp_seq=4 ttl=64 time=0.125 ms > 64 bytes from 172.16.1.11: icmp_seq=5 ttl=64 time=0.125 ms > 64 bytes from 172.16.1.11: icmp_seq=6 ttl=64 time=0.125 ms > 64 bytes from 172.16.1.11: icmp_seq=7 ttl=64 time=0.198 ms > 64 bytes from 172.16.1.11: icmp_seq=8 ttl=64 time=0.171 ms > 64 bytes from 172.16.1.11: icmp_seq=9 ttl=64 time=0.194 ms > 64 bytes from 172.16.1.11: icmp_seq=10 ttl=64 time=0.115 ms Out of curiosity, are you using connected mode or datagram mode for this? Also, are you using the inbuilt OS infiniband drivers, or Mellanox's OFED? (Or Intel/QLogic's equivalent if using their stuff) Asking because I haven't yet seen any real "best practise" stuff on ways to set this up for Gluster (yet). ;) Regards and best wishes, Justin Clift -- Open Source and Standards @ Red Hat twitter.com/realjustinclift ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
On 06/15/13 00:50, Stephan von Krawczynski wrote: Uh, you should throw away your GigE switch. Example: # ping 192.168.83.1 PING 192.168.83.1 (192.168.83.1) 56(84) bytes of data. 64 bytes from 192.168.83.1: icmp_seq=1 ttl=64 time=0.310 ms 64 bytes from 192.168.83.1: icmp_seq=2 ttl=64 time=0.199 ms 64 bytes from 192.168.83.1: icmp_seq=3 ttl=64 time=0.119 ms 64 bytes from 192.168.83.1: icmp_seq=4 ttl=64 time=0.115 ms What is the make and model of your GigE switch? I get: 114 packets transmitted, 114 received, 0% packet loss, time 113165ms rtt min/avg/max/mdev = 0.350/0.380/0.608/0.027 ms On a not loaded WS-C3560X-48. Though it might not be the switch. It could be the NIC on either side of the ping, Or anything up through the kernel, where the ping response is generated. Granted, my numbers are at home, between an Atom 330 and an AMD G-T56N, both with RealTek on motherboard NICs. AMD G-T56N <=> RealTek <=> WS-C3560X-48 <=> RealTek <=> Atom 330 So, now data from work: 48 packets transmitted, 48 received, 0% packet loss, time 47828ms rtt min/avg/max/mdev = 0.110/0.158/0.187/0.022 ms That is through a WS-C6513-E with a 2T supp card, then through the TOR WS-C3560X-48. So, I have lower latency with the ADDITION of the 6513 (not replacement, extra switch hop). Which means my NICs and up to Layer 7 (kernel) are the major players here. Work ping is between two identical HP DL360s (Xeon E5649, with Broadcom NetXtreme II GigE) Xeon E5649 <=> Broadcom <=> WS-C6513-E <=> WS-C3560X-48 <=> Broadcom <=> Xeon E5649 -- Mr. Flibble King of the Potato People ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
On Fri, 14 Jun 2013 14:35:26 -0700 Bryan Whitehead wrote: > GigE is slower. Here is ping from same boxes but using the 1GigE cards: > > [root@node0.cloud ~]# ping -c 10 10.100.0.11 > PING 10.100.0.11 (10.100.0.11) 56(84) bytes of data. > 64 bytes from 10.100.0.11: icmp_seq=1 ttl=64 time=0.628 ms > 64 bytes from 10.100.0.11: icmp_seq=2 ttl=64 time=0.283 ms > 64 bytes from 10.100.0.11: icmp_seq=3 ttl=64 time=0.307 ms > 64 bytes from 10.100.0.11: icmp_seq=4 ttl=64 time=0.275 ms > 64 bytes from 10.100.0.11: icmp_seq=5 ttl=64 time=0.313 ms > 64 bytes from 10.100.0.11: icmp_seq=6 ttl=64 time=0.278 ms > 64 bytes from 10.100.0.11: icmp_seq=7 ttl=64 time=0.309 ms > 64 bytes from 10.100.0.11: icmp_seq=8 ttl=64 time=0.197 ms > 64 bytes from 10.100.0.11: icmp_seq=9 ttl=64 time=0.267 ms > 64 bytes from 10.100.0.11: icmp_seq=10 ttl=64 time=0.187 ms > > --- 10.100.0.11 ping statistics --- > 10 packets transmitted, 10 received, 0% packet loss, time 9000ms > rtt min/avg/max/mdev = 0.187/0.304/0.628/0.116 ms > > Note: The Infiniband interfaces have a constant load of traffic from > glusterfs. The Nic cards comparatively have very little traffic. Uh, you should throw away your GigE switch. Example: # ping 192.168.83.1 PING 192.168.83.1 (192.168.83.1) 56(84) bytes of data. 64 bytes from 192.168.83.1: icmp_seq=1 ttl=64 time=0.310 ms 64 bytes from 192.168.83.1: icmp_seq=2 ttl=64 time=0.199 ms 64 bytes from 192.168.83.1: icmp_seq=3 ttl=64 time=0.119 ms 64 bytes from 192.168.83.1: icmp_seq=4 ttl=64 time=0.115 ms 64 bytes from 192.168.83.1: icmp_seq=5 ttl=64 time=0.099 ms 64 bytes from 192.168.83.1: icmp_seq=6 ttl=64 time=0.082 ms 64 bytes from 192.168.83.1: icmp_seq=7 ttl=64 time=0.091 ms 64 bytes from 192.168.83.1: icmp_seq=8 ttl=64 time=0.096 ms 64 bytes from 192.168.83.1: icmp_seq=9 ttl=64 time=0.097 ms 64 bytes from 192.168.83.1: icmp_seq=10 ttl=64 time=0.095 ms 64 bytes from 192.168.83.1: icmp_seq=11 ttl=64 time=0.097 ms 64 bytes from 192.168.83.1: icmp_seq=12 ttl=64 time=0.102 ms 64 bytes from 192.168.83.1: icmp_seq=13 ttl=64 time=0.103 ms 64 bytes from 192.168.83.1: icmp_seq=14 ttl=64 time=0.108 ms 64 bytes from 192.168.83.1: icmp_seq=15 ttl=64 time=0.098 ms 64 bytes from 192.168.83.1: icmp_seq=16 ttl=64 time=0.093 ms 64 bytes from 192.168.83.1: icmp_seq=17 ttl=64 time=0.099 ms 64 bytes from 192.168.83.1: icmp_seq=18 ttl=64 time=0.102 ms 64 bytes from 192.168.83.1: icmp_seq=19 ttl=64 time=0.092 ms 64 bytes from 192.168.83.1: icmp_seq=20 ttl=64 time=0.111 ms 64 bytes from 192.168.83.1: icmp_seq=21 ttl=64 time=0.112 ms 64 bytes from 192.168.83.1: icmp_seq=22 ttl=64 time=0.099 ms 64 bytes from 192.168.83.1: icmp_seq=23 ttl=64 time=0.092 ms 64 bytes from 192.168.83.1: icmp_seq=24 ttl=64 time=0.102 ms 64 bytes from 192.168.83.1: icmp_seq=25 ttl=64 time=0.108 ms ^C --- 192.168.83.1 ping statistics --- 25 packets transmitted, 25 received, 0% packet loss, time 23999ms rtt min/avg/max/mdev = 0.082/0.112/0.310/0.047 ms That is _loaded_. -- Regards, Stephan ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
GigE is slower. Here is ping from same boxes but using the 1GigE cards: [root@node0.cloud ~]# ping -c 10 10.100.0.11 PING 10.100.0.11 (10.100.0.11) 56(84) bytes of data. 64 bytes from 10.100.0.11: icmp_seq=1 ttl=64 time=0.628 ms 64 bytes from 10.100.0.11: icmp_seq=2 ttl=64 time=0.283 ms 64 bytes from 10.100.0.11: icmp_seq=3 ttl=64 time=0.307 ms 64 bytes from 10.100.0.11: icmp_seq=4 ttl=64 time=0.275 ms 64 bytes from 10.100.0.11: icmp_seq=5 ttl=64 time=0.313 ms 64 bytes from 10.100.0.11: icmp_seq=6 ttl=64 time=0.278 ms 64 bytes from 10.100.0.11: icmp_seq=7 ttl=64 time=0.309 ms 64 bytes from 10.100.0.11: icmp_seq=8 ttl=64 time=0.197 ms 64 bytes from 10.100.0.11: icmp_seq=9 ttl=64 time=0.267 ms 64 bytes from 10.100.0.11: icmp_seq=10 ttl=64 time=0.187 ms --- 10.100.0.11 ping statistics --- 10 packets transmitted, 10 received, 0% packet loss, time 9000ms rtt min/avg/max/mdev = 0.187/0.304/0.628/0.116 ms Note: The Infiniband interfaces have a constant load of traffic from glusterfs. The Nic cards comparatively have very little traffic. On Fri, Jun 14, 2013 at 12:40 PM, Stephan von Krawczynski wrote: > On Fri, 14 Jun 2013 12:13:53 -0700 > Bryan Whitehead wrote: > >> I'm using 40G Infiniband with IPoIB for gluster. Here are some ping >> times (from host 172.16.1.10): >> >> [root@node0.cloud ~]# ping -c 10 172.16.1.11 >> PING 172.16.1.11 (172.16.1.11) 56(84) bytes of data. >> 64 bytes from 172.16.1.11: icmp_seq=1 ttl=64 time=0.093 ms >> 64 bytes from 172.16.1.11: icmp_seq=2 ttl=64 time=0.113 ms >> 64 bytes from 172.16.1.11: icmp_seq=3 ttl=64 time=0.163 ms >> 64 bytes from 172.16.1.11: icmp_seq=4 ttl=64 time=0.125 ms >> 64 bytes from 172.16.1.11: icmp_seq=5 ttl=64 time=0.125 ms >> 64 bytes from 172.16.1.11: icmp_seq=6 ttl=64 time=0.125 ms >> 64 bytes from 172.16.1.11: icmp_seq=7 ttl=64 time=0.198 ms >> 64 bytes from 172.16.1.11: icmp_seq=8 ttl=64 time=0.171 ms >> 64 bytes from 172.16.1.11: icmp_seq=9 ttl=64 time=0.194 ms >> 64 bytes from 172.16.1.11: icmp_seq=10 ttl=64 time=0.115 ms >> >> --- 172.16.1.11 ping statistics --- >> 10 packets transmitted, 10 received, 0% packet loss, time 8999ms >> rtt min/avg/max/mdev = 0.093/0.142/0.198/0.035 ms > > What you like to say is that there is no significant difference compared to > GigE, right? > Anyone got a ping between two kvm-qemu virtio-net cards at hand? > > -- > Regards, > Stephan > ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
On Fri, 14 Jun 2013 12:13:53 -0700 Bryan Whitehead wrote: > I'm using 40G Infiniband with IPoIB for gluster. Here are some ping > times (from host 172.16.1.10): > > [root@node0.cloud ~]# ping -c 10 172.16.1.11 > PING 172.16.1.11 (172.16.1.11) 56(84) bytes of data. > 64 bytes from 172.16.1.11: icmp_seq=1 ttl=64 time=0.093 ms > 64 bytes from 172.16.1.11: icmp_seq=2 ttl=64 time=0.113 ms > 64 bytes from 172.16.1.11: icmp_seq=3 ttl=64 time=0.163 ms > 64 bytes from 172.16.1.11: icmp_seq=4 ttl=64 time=0.125 ms > 64 bytes from 172.16.1.11: icmp_seq=5 ttl=64 time=0.125 ms > 64 bytes from 172.16.1.11: icmp_seq=6 ttl=64 time=0.125 ms > 64 bytes from 172.16.1.11: icmp_seq=7 ttl=64 time=0.198 ms > 64 bytes from 172.16.1.11: icmp_seq=8 ttl=64 time=0.171 ms > 64 bytes from 172.16.1.11: icmp_seq=9 ttl=64 time=0.194 ms > 64 bytes from 172.16.1.11: icmp_seq=10 ttl=64 time=0.115 ms > > --- 172.16.1.11 ping statistics --- > 10 packets transmitted, 10 received, 0% packet loss, time 8999ms > rtt min/avg/max/mdev = 0.093/0.142/0.198/0.035 ms What you like to say is that there is no significant difference compared to GigE, right? Anyone got a ping between two kvm-qemu virtio-net cards at hand? -- Regards, Stephan ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] 40 gig ethernet
I'm using 40G Infiniband with IPoIB for gluster. Here are some ping times (from host 172.16.1.10): [root@node0.cloud ~]# ping -c 10 172.16.1.11 PING 172.16.1.11 (172.16.1.11) 56(84) bytes of data. 64 bytes from 172.16.1.11: icmp_seq=1 ttl=64 time=0.093 ms 64 bytes from 172.16.1.11: icmp_seq=2 ttl=64 time=0.113 ms 64 bytes from 172.16.1.11: icmp_seq=3 ttl=64 time=0.163 ms 64 bytes from 172.16.1.11: icmp_seq=4 ttl=64 time=0.125 ms 64 bytes from 172.16.1.11: icmp_seq=5 ttl=64 time=0.125 ms 64 bytes from 172.16.1.11: icmp_seq=6 ttl=64 time=0.125 ms 64 bytes from 172.16.1.11: icmp_seq=7 ttl=64 time=0.198 ms 64 bytes from 172.16.1.11: icmp_seq=8 ttl=64 time=0.171 ms 64 bytes from 172.16.1.11: icmp_seq=9 ttl=64 time=0.194 ms 64 bytes from 172.16.1.11: icmp_seq=10 ttl=64 time=0.115 ms --- 172.16.1.11 ping statistics --- 10 packets transmitted, 10 received, 0% packet loss, time 8999ms rtt min/avg/max/mdev = 0.093/0.142/0.198/0.035 ms On Fri, Jun 14, 2013 at 7:03 AM, Nathan Stratton wrote: > I have been playing around with Gluster on and off for the last 6 years or > so. Most of the things that have been keeping me from using it have been > related to latency. > > In the past I have been using 10 gig infiniband or 10 gig ethernet, recently > the price of 40 gig ethernet has fallen quite a bit with guys like Arista. > > My question is, is this worth it at all for something like Gluster? The port > to port latency looks impressive at under 4 microseconds, but I don't yet > know what total system to system latency would look like assuming QSPF+ > copper cables and linux stack. > > -- >><> > Nathan Stratton Founder, CTO > Exario Networks, Inc. > nathan at robotics.net nathan at > exarionetworks.com > http://www.robotics.net > http://www.exarionetworks.com/ > > Building the WebRTC solutions today that your customers will demand > tomorrow. > > ___ > Gluster-users mailing list > Gluster-users@gluster.org > http://supercolony.gluster.org/mailman/listinfo/gluster-users ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users