[Openstack] [SWIFT] When is it safe to remove data in /tmp?

Stephen Wood smwood4 at gmail.com
Wed Mar 5 20:54:31 UTC 2014


Does anyone have any thoughts on this?

I have a feeling it was because I rebalanced, pushed the shards, but then
rolled back to the original (before the rebalance) shards because I saw
that it was too aggressive.

During that time I saw errors like the following, which I believe is
responsible for the data in the /tmp directory. Does this data ever get
house-cleaned?

Mar  5 20:51:43 host1 object-server: ERROR Container update failed (saving
for async update later): 404 response from host3:6001/slot-7 (txn:
tx1df2ff494183469685552-0053178e5fdalswiftproxy04)



On Mon, Mar 3, 2014 at 4:41 PM, Stephen Wood <smwood4 at gmail.com> wrote:

> During a deployment of new container servers I noticed that my
> /srv/node/container/tmp directory ballooned. The files contained in the
> directory are sqlite DBs.
>
> The directory has been setting at 20GB for a few days and I'm starting to
> wonder if if the files (presumably set for a later async) are just
> orphaned. I'm not witnessing any container errors or other errors on the
> cluster.
>
> Is it possible they are orphans that can simply be removed?
>
> --
> Stephen Wood
> www.heystephenwood.com
>



-- 
Stephen Wood
www.heystephenwood.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack/attachments/20140305/e53dde63/attachment.html>


More information about the Openstack mailing list