[Bug 223238] Persistent iscsi underflow mismatch errors followed by apparent kernel panic

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=223238

--- Comment #7 from Jim D. <randomnoise058 at gmail.com> ---
On the target side, I know that the Synology SAN/NAS has some disk monitoring.
There are scheduled smart tests, but they are weeks to months apart. All
Synology administrative console smartctl resuls show "Normal" for the affected
target HDD. There are also several Synology applications running related to
"cloud services" (thumbd,syncd,cloud-cleand,storaged). There is also an
ecryptfs-kthread task and a synoindexd task. There have been a couple of
Synology SAN/NAS firwmware updates over the past month or so.

On the initiator side - I am occassionally running Zabbix to query hard drives
(client script: ls,dd,fsck,fdisk,iostat,smartctl,diskutil,system_profiler)
every 10-15 minutes. On some targets I also run ZFS queries and iSCSI
(lsscsi,iscsiadm,iscsictl) queries. But nothing that does presistent monitoring
(like every few seconds or non-stop). I have at times 8-10 client hosts
(physical and virtual) monitored by either a physical Zabbix or a virtual
Zabbix, all of which have iSCSI targets.Right now I have 12 targets with 19
LUNs defined attached to 6 different *nix OS's/versions.

I am unable to see any more screen output when the FBSD VirtualBox machine
locks up. Neither the VBox console nor the SSH terminal are responsive. I might
be able to run a trace to a file, but don't know if that will capture what you
are interested in - the data might never get written to local (VBox) disk even
if there was something worth seeing. When I look at locked-up-client-host
messages, there is nothing but "last timed entry" and the next startup entry.

I have enabled the /etc/syslog.conf configuration for the "console log".
Perhaps that will capture some additional information. Past experience has
shown, however, that open log files during a system crash/lock up may not
receive all of the possible data in the log file.

-- 
You are receiving this mail because:
You are the assignee for the bug.