Hello, I would like to test 6.2-STABLE's ability to handle UFS2, quotas, and snapshots. In particular, I would like to make sure the deadlock issue is resolved, (http://people.freebsd.org/~kib/quotagiant/quotas-20060428-1252.patch) In addition, it appears that compiling support for quotas in the kernel will force the Giant lock on UFS and I want to make sure I won't have a performance regression on the filesystem. Can anyone recommend a testing regime? I have a -STABLE production machine in mind that needs quotas enabled, is running sysutils/freebsd-snapshot, and has several large filesystems (100GB+). I'm not sure I want to "test" on this machine, however. Thanks for your help. -- Regards, Doug
On Tue, Mar 13, 2007 at 09:50:54AM -0500, Doug Poland wrote:> Hello, > > I would like to test 6.2-STABLE's ability to handle UFS2, quotas, and > snapshots. In particular, I would like to make sure the deadlock > issue is resolved, > (http://people.freebsd.org/~kib/quotagiant/quotas-20060428-1252.patch)This patch is for Giant removal around UFS with QUOTAS, is outdated and contains known bugs. Please, wait several days until I finally commit stuff into the HEAD. After that, I most likely will backport the patch.> > In addition, it appears that compiling support for quotas in the > kernel will force the Giant lock on UFS and I want to make sure I > won't have a performance regression on the filesystem.See above.> > Can anyone recommend a testing regime? I have a -STABLE production > machine in mind that needs quotas enabled, is running > sysutils/freebsd-snapshot, and has several large filesystems (100GB+). > I'm not sure I want to "test" on this machine, however. > > Thanks for your help. >-------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 187 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/freebsd-stable/attachments/20070313/71ca47ca/attachment.pgp
On Tue, Mar 13, 2007 at 09:50:54AM -0500 I heard the voice of Doug Poland, and lo! it spake thus:> > In addition, it appears that compiling support for quotas in the > kernel will force the Giant lock on UFS and I want to make sure I > won't have a performance regression on the filesystem.I can tell you that the impact will range from negligable to significant, depending on the particular workload. I have a RELENG_6 box that, for hysterical raisins, has the ports tree on a partition with quotas, and a large 'portsnap update' gets VERY jerky. -- Matthew Fuller (MF4839) | fullermd@over-yonder.net Systems/Network Administrator | http://www.over-yonder.net/~fullermd/ On the Internet, nobody can hear you scream.