ZFS out of swap space

Xin Li delphij at delphij.net
Mon Mar 30 18:20:00 UTC 2015


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

On 03/30/15 01:32, armonia wrote:
> Yes, my mistake was probably that I have included data
> deduplication to see how it works, but not turned it off at the
> right time. In this case, the machine memory of 4 GB ....
> 
> async_destroy - too enabled.
> 
> That is the conclusion I have deduplication disabled.

Hrm, usually async_destroy should be enough to protect against this
situation.

Can you try if setting the sysctl variable vfs.zfs.free_max_blocks to
a limited number, like, 100,000, would make the pool import properly?
 You may have run out of memory because of too much data being free'ed
in one transaction group?

(I would recommend doing this after importing the pool read-only and
copy your data off, though).

> How to import a pool of read-only?

zpool import -o readonly poolname.

> Thank you for your response.
> 
> zpool get all zroot NAME   PROPERTY                       VALUE
> SOURCE zroot  size                           230G
> - zroot  capacity                       24%
> - zroot  altroot                        -
> default zroot  health                         ONLINE
> - zroot  guid                           1229884058434432944
> default zroot  version                        -
> default zroot  bootfs                         zroot
> local zroot  delegation                     on
> default zroot  autoreplace                    on
> local zroot  cachefile                      -
> default zroot  failmode                       wait
> default zroot  listsnapshots                  on
> local zroot  autoexpand                     off
> default zroot  dedupditto                     0
> default zroot  dedupratio                     1.02x
> - zroot  free                           174G
> - zroot  allocated                      56.1G
> - zroot  readonly                       off
> - zroot  comment                        ZFS
> local zroot  expandsize                     0
> - zroot  freeing                        0
> default zroot  feature at async_destroy          enabled
> local zroot  feature at empty_bpobj            active
> local zroot  feature at lz4_compress           active
> local zroot  feature at multi_vdev_crash_dump  enabled
> local zroot  feature at spacemap_histogram     active
> local zroot  feature at enabled_txg            active
> local zroot  feature at hole_birth             active
> local zroot  feature at extensible_dataset     enabled
> local zroot  feature at bookmarks              enabled
> local zroot  feature at filesystem_limits      enabled
> local
> 
> 
> ???????????, 30 ????? 2015, 0:36 -07:00 ?? Xin Li
> <delphij at delphij.net>:
> 
> 
> 
> On 3/27/15 05:26, armonia wrote:
>> After importing I press ctrl + t and here's the conclusion:
> 
>> load: 0.20 cmd: zpool 725 [tx->tx_sync_done_cv] 32.50r 0.00y
>> 5.59s 0% 6432k
> 
> Have you ever enabled e.g. dedup on certain dataset(s) and have a
> lot of files, and the pool don't have 'async destroy' feature
> enabled? In that case the fastest way to recover, if this took too
> long, would probably import the pool read-only and copy data to
> another pool.
> 
> On -CURRENT you can use dtrace -qn 'zfs-dbgmsg{printf("%s\n", 
> stringof(arg0))}' to see more verbose information, they are not
> always helpful but will give you better idea on what is going on
> under the hood .
> 
> Cheers, _______________________________________________ 
> freebsd-stable at freebsd.org 
> </compose?To=freebsd%2dstable at freebsd.org> mailing list 
> http://lists.freebsd.org/mailman/listinfo/freebsd-stable To
> unsubscribe, send any mail to 
> "freebsd-stable-unsubscribe at freebsd.org 
> </compose?To=freebsd%2dstable%2dunsubscribe at freebsd.org>"
> 
> 
> 
> --

- -- 
Xin LI <delphij at delphij.net>    https://www.delphij.net/
FreeBSD - The Power to Serve!           Live free or die
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.1.2 (FreeBSD)
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=m5xz
-----END PGP SIGNATURE-----


More information about the freebsd-stable mailing list