Bug#878441: lvm2: commands that flush cache do not complete
Michael F. Lamb
mike at datagrok.org
Fri Oct 13 18:14:09 UTC 2017
Package: lvm2
Version: 2.02.173-1
Severity: normal
I have an lv with a writeback cache:
$ sudo lvs -a -o+cachemode
LV VG Attr LSize Pool Origin Data% Meta% Move Log Cpy%Sync Convert CacheMode
...
home linux Cwi-a-C--- 130.00g [homecache] [home_corig] 99.94 1.72 90.77 writeback
[home_corig] linux owi-aoC--- 130.00g
[home_corig] linux owi-aoC--- 130.00g
[home_corig] linux owi-aoC--- 130.00g
[homecache] linux Cwi---C--- 104.16g 99.94 1.72 90.77 writeback
[homecache_cdata] linux Cwi-ao---- 104.16g
[homecache_cmeta] linux ewi-ao---- 20.00m
After a hard-reset and an fsck which performed several repairs, I
attempted to change the cachemode from 'writeback' to 'writethrough'.
The progress reaches a certain point but then seems to loop forever.
$ sudo lvchange --cachemode writethrough linux/home
Flushing 24191 blocks for cache linux/home.
Flushing 24190 blocks for cache linux/home.
Flushing 24189 blocks for cache linux/home.
Flushing 24189 blocks for cache linux/home.
Flushing 24189 blocks for cache linux/home.
[ ... repeating for hours ... ]
Flushing 24189 blocks for cache linux/home.
^C Interrupted...
Flushing of linux/home aborted.
Then I attempted to remove the cache altogether, with the same result.
$ sudo lvconvert --uncache linux/home
Flushing 24189 blocks for cache linux/home.
[ ... repeating for hours ... ]
Flushing 24189 blocks for cache linux/home.
^C Interrupted...
Flushing of linux/home aborted.
This issue seems like it might be related to upstream
https://bugzilla.redhat.com/show_bug.cgi?id=1276722 but a comment on
that bug states it was dealt with back in version 2.02.133. (I'm on
2.02.173.)
Note: during this issue, I also experienced bug #857142, so had to
perform the fsck after booting from a debian live usb drive. The
lvchange and lvconvert commands however were performed from this system.
I ran SMART extended diagnostics (including surface scan) on the drives
hosting the PVs for both the cache and home LVs and it reported no
problems. I can't find any I/O errors in my system logs either.
I don't know what initially caused the crash requiring me to hard-reset.
I think it was related to switching my xorg console while using
proprietary video drivers.
I copied (with tar) the data over to a new non-cached lv for my /home,
but I still have the old one around if you would like additional
information about it.
-- System Information:
Debian Release: buster/sid
APT prefers testing
APT policy: (990, 'testing'), (500, 'stable'), (90, 'unstable')
Architecture: amd64 (x86_64)
Foreign Architectures: i386
Kernel: Linux 4.12.0-2-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en_US.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)
Versions of packages lvm2 depends on:
ii dmeventd 2:1.02.142-1
ii dmsetup 2:1.02.142-1
ii init-system-helpers 1.49
ii libblkid1 2.29.2-5+b1
ii libc6 2.24-17
ii libdevmapper-event1.02.1 2:1.02.142-1
ii libdevmapper1.02.1 2:1.02.142-1
ii liblvm2app2.2 2.02.173-1
ii libreadline5 5.2+dfsg-3+b1
ii libudev1 234-3
ii lsb-base 9.20170808
lvm2 recommends no packages.
Versions of packages lvm2 suggests:
ii thin-provisioning-tools 0.6.1-4+b1
-- no debconf information
More information about the pkg-lvm-maintainers
mailing list