Jesus Cea
2011-Jul-25 04:17 UTC
[zfs-discuss] Unexpected "out of space" when creating snapshots
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I am creating a recursive snapshot in my ZPOOL and I am getting an unexpected error: """ [root at stargate-host /]# ./z-snapshotZFS 20110725-05:56 cannot create snapshot ''datos/swap at 20110725-05:56'': out of space no snapshots were created [root at stargate-host /]# zpool list NAME SIZE ALLOC FREE CAP HEALTH ALTROOT datos 688G 629G 59.1G 91% ONLINE - [root at stargate-host /]# zfs get all datos/swap NAME PROPERTY VALUE SOURCE datos/swap type volume - datos/swap creation Thu Jul 16 17:15 2009 - datos/swap used 32G - datos/swap available 30.2G - datos/swap referenced 18.3G - datos/swap compressratio 1.31x - datos/swap reservation none default datos/swap volsize 32G local datos/swap volblocksize 4K - datos/swap checksum on default datos/swap compression on inherited from datos datos/swap readonly off default datos/swap shareiscsi off default datos/swap copies 1 default datos/swap refreservation 32G local datos/swap primarycache none local datos/swap secondarycache none local datos/swap usedbysnapshots 0 - datos/swap usedbydataset 0 - datos/swap usedbychildren 0 - datos/swap usedbyrefreservation 0 - datos/swap logbias latency default """ 32GB of swap is overkill for most people, but my data is huge but the working set fits memory. "z-snapshotZFS" is a script doing a recursive ZFS snapshot of all my datasets and then deleting some of my snapshots like "dump" or "swap", email spool, etc. Any idea?. I have almost 60GB free in my pool... Even with the huge ZPOOL write reservation, I have: """ [root at stargate-host /]# zfs list datos NAME USED AVAIL REFER MOUNTPOINT datos 661G 16.6G 170M /datos """ 16.6GB free. Or is the 32GB reservation in "datos/swap" dataset hurting me here?. (BTW, wasting 50GB on the disk, or 7% of it is not nice. I know that that is for performance, but still...). If that is the case, how could I delete the 18GB references in "datos/swap"?. I am using only a bit of swap at this moment... """ [root at stargate-host /]# swap -s total: 1778312k bytes allocated + 261088k reserved = 2039400k used, 33412568k available """ 2GB, not 18GB... Should Solaris "trim" the swap space after boot/when it is swapped in?. I am using Solaris 10 Update 9. Thanks for your time. PS: Alternatively, could I do a recursive snapshotting of all my datasets skipping over a few of them?. Now I do a full recursive snapshot, and then delete specific snapshots I don''t want to have, like the swap. - -- Jesus Cea Avion _/_/ _/_/_/ _/_/_/ jcea at jcea.es - http://www.jcea.es/ _/_/ _/_/ _/_/ _/_/ _/_/ jabber / xmpp:jcea at jabber.org _/_/ _/_/ _/_/_/_/_/ . _/_/ _/_/ _/_/ _/_/ _/_/ "Things are not so easy" _/_/ _/_/ _/_/ _/_/ _/_/ _/_/ "My name is Dump, Core Dump" _/_/_/ _/_/_/ _/_/ _/_/ "El amor es poner tu felicidad en la felicidad de otro" - Leibniz -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iQCVAwUBTizuP5lgi5GaxT1NAQJEzQP9Ht9l8r+z2LeFOXI23m5RdCn+lnWnL9sD 4DPOr7fIn8noJ5/YW7+uusFKaOi51BzQMEtvgNDNN30SCEOZo5ZaSjU2EKJnxvdd 09lUlv1VUbFe9mmZPaLutEx4CNwbBjJBcyeviUbTlqJQWpjli4zASxDKjcnkuUY+ 9HLwujW2iCc=8dDa -----END PGP SIGNATURE-----
Ian Collins
2011-Jul-25 05:15 UTC
[zfs-discuss] Unexpected "out of space" when creating snapshots
On 07/25/11 04:17 PM, Jesus Cea wrote:> I am creating a recursive snapshot in my ZPOOL and I am getting an > unexpected error: > > """ > [root at stargate-host /]# ./z-snapshotZFS 20110725-05:56 > cannot create snapshot ''datos/swap at 20110725-05:56'': out of space > no snapshots were createdSwap can have a huge churn, how often do you snapshot? I''ve seen swap snapshots almost as big as the volume! Put everything you want to snapshot under a single root filesystem ad do a recursive snapshot on that. -- Ian.