Re: [HPDD-discuss] Lustre 2.6.0 vs 2.5.3 on production systems

Bob, About performance, as well. In general, 2.6/2.7 client performance is significantly better than 2.5 (server performance is largely the same). There are certain specific cases with particular hardware configurations and workloads that are worse, but in general, I'd say improved client performance is perhaps the best reason to choose 2.6 over 2.5.3. (Certainly 2.6 is going to be a bit buggier than 2.5.3, which as the maintenance branch has had a significant amount of work done on it since 2.6 was released.) - Patrick On 03/13/2015 09:28 AM, Bob Ball wrote: > Thank you, Peter. That is helpful information. We are currently leaning > towards 2.6, but if the upgrade path from there to 2.7 is simple, and can > work with OST/MDT created under 2.6, then that may be the path that we'll > follow. We need to bring the new system online quite soon. > > Regards, > bob > > On 3/13/2015 10:20 AM, Jones, Peter A wrote: > >> Hi Bob >> >> I know that you are really looking for responses directly from sites >> running the releases in production but a couple of things that you might >> be helpful for you to know: >> >> 1) The recent OpenSFS survey 10 out of the 89 respondents said that they >> were using 2.6 in production >> 2) Lustre 2.7 is expected to be released quite soon and will be more >> current in kernel version, zfs version and bug fixes. >> >> Regards >> >> Peter >> >> On 3/13/15, 6:49 AM, "Bob Ball" <ball(a)umich.edu&gt; wrote: >> >> Hi, >>> >>> We are about to set up a new Lustre system here, and we are trying to >>> decide between using 2.5.3 or 2.6.0 . Files in our older 2.1.6 system >>> will simply be migrated over, and the old system tossed down the tubes. >>> >>> Is anyone using 2.6.0 in production? I found this on some 2.6.0 >>> performance data, and it looks rather discouraging. >>> https://jira.hpdd.intel.com/browse/LU-4841?filter=10828 >>> I have also noticed that the 2.5.3 server rpms actually use a slightly >>> newer kernel than the 2.6.0 set. By and large, we will use zfs >>> underlying the OST. >>> >>> So, I will gladly accept any advice on which way we should turn as we >>> set this up. >>> >>> Thanks, >>> bob >>> >>> _______________________________________________ >>> HPDD-discuss mailing list >>> HPDD-discuss(a)lists.01.org >>> https://lists.01.org/mailman/listinfo/hpdd-discuss >>> >> >> > _______________________________________________ > HPDD-discuss mailing list > HPDD-discuss(a)lists.01.org > https://lists.01.org/mailman/listinfo/hpdd-discuss > _______________________________________________ HPDD-discuss mailing list HPDD-discuss(a)lists.01.org https://lists.01.org/mailman/listinfo/hpdd-discuss