Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
Hi Jan ! This is a tough question. It depends on many parameters, starting with the size of the requested IOs. There are a couple of good articles on the net talking about that, including these : - 2 from Scott LOWE, wich are fairly clear and interesting from the IOPS estimation methodology standpoint : http://www.techrepublic.com/blog/datacenter/calculate-iops-in-a-storage-arra y/2182 http://www.techrepublic.com/blog/datacenter/storage-array-capacity-performan ce-vs-cost/2239 - This one from HP gives some interesting numeric values about the drives they supply in their servers : http://forums13.itrc.hp.com/service/forums/questionanswer.do?admit=109447627 +1305886571915+28353475 <http://forums13.itrc.hp.com/service/forums/questionanswer.do?admit=10944762 7+1305886571915+28353475&threadId=1122630> &threadId=1122630 - There’s also this one from Duncan EPPING, that provides a few additionnal informations : http://www.yellow-bricks.com/2009/12/23/iops/ But unfortunately, the real and immediate answer to your question is “It depends”! J Hope this helps, Best regards, Pascal. De : drbd-user-bounces at lists.linbit.com [mailto:drbd-user-bounces at lists.linbit.com] De la part de Jan.Dreyer at bertelsmann.de Envoyé : vendredi 20 mai 2011 10:11 À : drbd-user at lists.linbit.com Objet : [DRBD-user] Performance question Hi, I have a question that is not directly related to drbd, but more to performance of drbd on HW-Raid. This is our configuration: 2 HP machines with ubuntu-Linux (LTS) configured HW Raid5 on 4 disks (DG146BB976 = 146.8GB 10k SAS) on each of the two machines LVM on that HW-RAID with 3 logical volumes (system, swap, data) DRBD on data-volume (ext4) over dedicated GiB-interface Now the question: what performance (iops) can I expect from that configuration? I don’t really found useful information in the net, so I ask here. The drbd-interface gives about 40 tps (iostat -dk drbd1 5), which seems low to me. Also suspicious: %util is many times more than 98, and await is nearly always greater than svctm. These hints would allow to suspect an error, if I had only one disk – I have no idea which values to expect on my Raid5-configuration. Has anyone a hint or a guidance to how to calculate expectable values? Greetings Jan Dreyer -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20110520/e8561b0f/attachment.htm>