On Sat, Dec 13, 2014 at 12:20 AM, Michael Liu <[email protected]> wrote:
>
> I have a cluster of 2 nodes and one node was going out of memory
> performing compaction. So, I did a failover, and now the remaining server
> is unresponsive to my nodejs client. I have access to all my views and
> documents through http rest, so I'm guessing something is wrong with the
> way smart clients connect and make requests.
>
> In the couchdb log is full of these timeouts:
>
> [couchdb:error,2014-12-13T8:10:50.001,[email protected]:<0.468.0>:
> couch_log:error:44]dcp client (loop, mapreduce_view: loop 
> _design/sync_housekeeping
> (prod/main)): Obtaining mutation from server timed out after 60.0 seconds
> [RequestId 123, PartId 532, StartSeq 23, EndSeq 25]. Waiting...
>
> In the web console I see a popup that just says Indexing all my design
> docs/views, with the progress bar at 0.
>
> I was thinking to perform a backup/restore, but cbbbackup won't complete
> either:
> couchbase@union-1:~$ bin/cbbackup -v http://localhost:8091 backups/
> 2014-12-13 08:04:49,305: mt cbbackup...
> 2014-12-13 08:04:49,305: mt  source : http://localhost:8091
> 2014-12-13 08:04:49,305: mt  sink   : backups/
> 2014-12-13 08:04:49,305: mt  opts   : {'username': None, 'verbose': 1,
> 'dry_run': False, 'extra': {'max_retry': 10.0, 'rehash': 0.0, 'data_only':
> 0.0, 'uncompress': 0.0, 'nmv_retry': 1.0, 'conflict_resolve': 1.0,
> 'cbb_max_mb': 100000.0, 'mcd_compatible': 1.0, 'try_xwm': 1.0,
> 'backoff_cap': 10.0, 'batch_max_bytes': 400000.0, 'report_full': 2000.0,
> 'seqno': 0.0, 'batch_max_size': 1000.0, 'report': 5.0, 'design_doc_only':
> 0.0, 'recv_min_bytes': 4096.0}, 'single_node': False, 'vbucket_list': None
> , 'threads': 4, 'mode': 'full', 'key': None, 'password': None, 'id': None,
> 'bucket_source': None}
> 2014-12-13 08:04:49,320: mt bucket: loop
>
>
>
> ^Cinterrupted.######   ] 83.2% (6474/estimated 7780 msgs)
>
>
> Any help appreciated, at a loss here...
>

This looks like some bug within memcached. File bug at
http://www.couchbase.com/issues/secure/Dashboard.jspa. Attach
cbcollectinfos and in this case it might be useful to grab backtraces of
all threads in memcached. E.g.

# gdb -p <pid of your memcached>

> thread app all bt

> quit


>

-- 
You received this message because you are subscribed to the Google Groups 
"Couchbase" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to