Date: Wed, 20 Nov 2019 08:11:06 +0000 From: bugzilla-noreply@freebsd.org To: bugs@FreeBSD.org Subject: [Bug 241980] panic: I/O to pool appears to be hung on vdev Message-ID: <bug-241980-227-CQ4OGuH2YE@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-241980-227@https.bugs.freebsd.org/bugzilla/> References: <bug-241980-227@https.bugs.freebsd.org/bugzilla/>
next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D241980 --- Comment #19 from Andriy Gapon <avg@FreeBSD.org> --- (In reply to Eugene Grosbein from comment #18) There is no reason why deadman_enabled had to be read-only. But the code ne= eds to be careful to stop and restart the callout when the deadman logic is disabled and enabled. Maybe we can have the callout always running, then we need to check the flag in the callout callback. Also, I think that there is a bug in the early warning code. I think that = the warning timeout is 8 times longer that the regular timeout. The best way to debug the problem would be to get a crash dump. Maybe you could try to add a dedicated dump disk and connect it to a differ= ent controller, etc. But if you can't get that to work, then I suggest that you add more useful information to the early warning. You can print a pointer to zio and a poi= nter to vdev. Then you'd be able to use kgdb on a live kernel and you would hav= e a place to start. Just my 2 cents as they say. --=20 You are receiving this mail because: You are the assignee for the bug.=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-241980-227-CQ4OGuH2YE>