Re: NVME aborting outstanding i/o and controller resets

2019-04-15 Thread Patrick M. Hausen
Some updates: https://www.ixsystems.com/community/threads/nvme-problems-are-there-nightlies-based-on-12-stable-already.75685 https://jira.ixsystems.com/browse/NAS-101427 Kind regards, Patrick -- punkt.de GmbH Internet - Dienstleistungen - Beratung Kaiserallee 13a

Re: NVME aborting outstanding i/o and controller resets

2019-04-15 Thread Patrick M. Hausen
Hi! > Am 15.04.2019 um 10:51 schrieb Patrick M. Hausen : > Now, RELENG_12 kernel, 11.2-RELEASE userland: > > root@hurz:/var/tmp # uname -a > FreeBSD hurz 12.0-STABLE FreeBSD 12.0-STABLE r346220 GENERIC amd64 > root@hurz:/var/tmp # dd if=/dev/urandom of=hurz bs=10m > > Result: > > no problems,

Re: NVME aborting outstanding i/o and controller resets

2019-04-15 Thread Patrick M. Hausen
> Am 15.04.2019 um 08:46 schrieb Patrick M. Hausen : > So I’ll test RELENG_12 next. If that works, I can probably craft > a FreeNAS 11.2 installation with a 12 kernel. I would be hesitating to run > HEAD in production, though. root@hurz:/var/tmp # uname -a FreeBSD hurz 11.2-RELEASE FreeBSD 11.2-RE

Re: NVME aborting outstanding i/o and controller resets

2019-04-14 Thread Patrick M. Hausen
Hi! > Am 14.04.2019 um 23:33 schrieb Patrick M. Hausen : > Since the system runs well with RELENG_11 and only 4 drives > and there is this question about the cabling and shared resources > I will try to set up a system with 5 drives, each of them *without* > another one in a „pair“ sharing the sam

Re: NVME aborting outstanding i/o and controller resets

2019-04-14 Thread Patrick M. Hausen
Alright ... > Am 13.04.2019 um 02:37 schrieb Warner Losh : > > There's been some minor improvements in -current here. Any chance you could > > experimentally try that with this test? You won't get as many I/O abort > > errors (since we don't print those), and we have a few more workarounds f

Re: NVME aborting outstanding i/o and controller resets

2019-04-12 Thread Warner Losh
On Fri, Apr 12, 2019, 1:22 PM Patrick M. Hausen wrote: > Hi Warner, > > thanks for taking the time again … > > > OK. This means that whatever I/O workload we've done has caused the NVME > card to stop responding for 30s, so we reset it. > > I figured as much ;-) > > > So it's an intel card. > > Y

Re: NVME aborting outstanding i/o and controller resets

2019-04-12 Thread Patrick M. Hausen
Hi Warner, thanks for taking the time again … > OK. This means that whatever I/O workload we've done has caused the NVME card > to stop responding for 30s, so we reset it. I figured as much ;-) > So it's an intel card. Yes - I already added this info several times. 6 of them, 2.5“ NVME „disk

Re: NVME aborting outstanding i/o and controller resets

2019-04-12 Thread Warner Losh
On Fri, Apr 12, 2019 at 6:00 AM Patrick M. Hausen wrote: > Hi all, > > my problems seem not to be TRIM related after all … and I can now > quickly reproduce it. > > = > root@freenas01[~]# sysctl vfs.zfs.trim.enabled > vfs.zfs.trim.enabled: 0 > = > root@freenas01[~]# cd /mnt/zfs > root@fre

Re: NVME aborting outstanding i/o and controller resets

2019-04-12 Thread Patrick M. Hausen
Hi all, my problems seem not to be TRIM related after all … and I can now quickly reproduce it. = root@freenas01[~]# sysctl vfs.zfs.trim.enabled vfs.zfs.trim.enabled: 0 = root@freenas01[~]# cd /mnt/zfs root@freenas01[/mnt/zfs]# dd if=/dev/urandom of=hurz bs=10m ^C — system freezes tempora