mjmac@clusterfs.com
2007-Jan-04 15:34 UTC
[Lustre-devel] [Bug 11499] Performance regression from 1.4.7 -> 1.4.8
Please don''t reply to lustre-devel. Instead, comment in Bugzilla by using the following link: https://bugzilla.lustre.org/show_bug.cgi?id=11499 Removing privacy bits.
sss@cray.com
2007-Jan-09 09:37 UTC
[Lustre-devel] [Bug 11499] Performance regression from 1.4.7 -> 1.4.8
Please don''t reply to lustre-devel. Instead, comment in Bugzilla by using the following link: https://bugzilla.lustre.org/show_bug.cgi?id=11499 Is this blocking 1.4.9? If so, shouldn''t it be higher than S3/P3?
mjmac@clusterfs.com
2007-Jan-09 10:46 UTC
[Lustre-devel] [Bug 11499] Performance regression from 1.4.7 -> 1.4.8
Please don''t reply to lustre-devel. Instead, comment in Bugzilla by using the following link: https://bugzilla.lustre.org/show_bug.cgi?id=11499 Ran through IOR again, this time with a reformat before each run. Excessive, probably, but rules out any strangeness with on-disk structures. 1.4.8 still fails across the board. There were two slow runs on 1.4.7; I''m not sure what those can be attributed to. No outright timeouts as we are seeing with 1.4.8. I''ll queue up a v1_4_8_RC8 run to see if that branch really is somehow performing better than v1_4_8, even though I can''t see how. v1_4_7 Please don''t reply to lustre-devel. Instead, comment in Bugzilla by using the following link: https://buffalo.lustre.org:8443/display_results.pl?timespan=-14&branch=v1_4_7&name=ior&totalstatus=&testgroup=&host=boulder&arch=&networktype=&filesystem=&status=&ordered_by=created&direction=DESC&comparison_field=&comparison=is&comparison_text v1_4_8 Please don''t reply to lustre-devel. Instead, comment in Bugzilla by using the following link: https://buffalo.lustre.org:8443/display_results.pl?timespan=-14&branch=v1_4_8&name=ior&totalstatus=&testgroup=&host=boulder&arch=&networktype=&filesystem=&status=&ordered_by=created&direction=DESC&comparison_field=&comparison=is&comparison_text=