Date: Mon, 26 Aug 2019 20:44:59 +0000 From: bugzilla-noreply@freebsd.org To: bugs@FreeBSD.org Subject: [Bug 234576] hastd exits ungracefully Message-ID: <bug-234576-227-UdYTuvRA8e@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-234576-227@https.bugs.freebsd.org/bugzilla/> References: <bug-234576-227@https.bugs.freebsd.org/bugzilla/>
index | next in thread | previous in thread | raw e-mail
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=234576 Michael Tratz <michael@esosoft.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |michael@esosoft.com --- Comment #6 from Michael Tratz <michael@esosoft.com> --- I have the same issue with hastd exiting immediately since I upgraded to FreeBSD 11.3. I compared zfs sysctls with 11.2 and 11.3 and I found the culprit. All you have to do is set via sysctl or sysctl.conf: vfs.zfs.vdev.aggregation_limit=131072 In FreeBSD 11.3 and 12.x it was increased to 1MB. As soon as do that the pool imports without any problems using hastd. No patches needed. Hastd has been working stable for years for me and it seems to continue using this setting. The FreeBSD revision which introduced this bug is: base r334844 Line 178 in sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_queue.c https://svnweb.freebsd.org/base/head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_queue.c?r1=334844&r2=334843&pathrev=334844 On my spinning rust systems I don't see any speed difference for scrubs with the setting set to 128KB or 1MB. -- You are receiving this mail because: You are the assignee for the bug.help
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-234576-227-UdYTuvRA8e>