Hi Ihara, > Frank, > > > tune2fs -O ^has_journal<ost-device> on all OSSes for all all OSTs > > What is a reason of removing the journal feature? You should have the journal > even whether you want to enable > async_journal or not. > That was actually the question. Many thanks for the clarification.
> Also, as far as I could see, in general, without enabling async journal, we > could see better performance on using > external journals rather than the internal journal. > However, once async_journal is enabled, we could see same (in some case, > better) performance even the > internal journal is used. > Well, the reason why this feature is so interessting is described in paper: http://www.usenix.org/events/fast10/tech/full_papers/oral.pdf It confirms your first statement. Anyway async journal feature seems to improve any 'low-level' HW solution. Many thanks for the comments. -Frank > Thanks > Ihara > > (7/13/10 9:53 PM), Frank Heckes wrote: > > Hi all, > > > > we use SLES 11 and Lustre 1.8.1.1 + patches and like convert a lustre FS > > using external journals to one with async journals enabled. > > Question is whether the procedure: > > > > umount<filesystem> on all clients > > umount<osts> on all OSSes > > e2fsck<ost-device> on all OSSes for all all OSTs > > tune2fs -O ^has_journal<ost-device> on all OSSes for all all OSTs > > lctl set_param obdfilter.*.sync_journal=0 on all OSSes > > mount<osts> on all OSSes > > mount<filesystem> on all clients > > > > is correct to do the job? (I hope it isn't neccessary to recreate a FS > > from scratch.) Many thanks in advance. > > > > Cheers > > > > -Frank Heckes > > > > P.S.: 1.8.1.1 still contains some bugs which have been fixed in 1.8.3. > > Described setup is for test purpose only, but the procedure shall be > > used in the final environment (using lustre 1.8.3), too. > > > > > > > > ------------------------------------------------------------------------------------------------ > > ------------------------------------------------------------------------------------------------ > > Forschungszentrum Juelich GmbH > > 52425 Juelich > > Sitz der Gesellschaft: Juelich > > Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 > > Vorsitzender des Aufsichtsrats: MinDirig Dr. Karl Eugen Huthmacher > > Geschaeftsfuehrung: Prof. Dr. Achim Bachem (Vorsitzender), > > Dr. Ulrich Krafft (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt, > > Prof. Dr. Sebastian M. Schmidt > > ------------------------------------------------------------------------------------------------ > > ------------------------------------------------------------------------------------------------ > > _______________________________________________ > > Lustre-discuss mailing list > > [email protected] > > http://lists.lustre.org/mailman/listinfo/lustre-discuss > _______________________________________________ > Lustre-discuss mailing list > [email protected] > http://lists.lustre.org/mailman/listinfo/lustre-discuss _______________________________________________ Lustre-discuss mailing list [email protected] http://lists.lustre.org/mailman/listinfo/lustre-discuss
