CouchDB 2.0 cluster problem when first node is down

366 views
Skip to first unread message

Oleg Cohen

unread,
May 6, 2016, 2:18:36 PM5/6/16
to us...@couchdb.apache.org
Greetings,

We ran into an issue when testing CouchDB 2.0 clustering. We ran a 2-node
cluster using the dev/run -n 2 command.

If we bring node2 down, all on node1 is still working fine. When we bring
down node1 using remsh and issuing init:stop(). command, the databases on
node2 are no longer readable.

We test this by trying to read the _users database using the following
command:

curl -X GET "http://127.0.0.1:25984/_users" --user admin:xxxxxxx

The get the following error:

{"error":"nodedown","reason":"progress not possible"}

If node1 is restarted, the problem goes away.

We experienced the same issue running a 3-node cluster across 3 different
servers.

Wondering if anyone ran into the same issue and if there is a workaround or
a way to fix the issue.

Thank you,
Oleg


--
*Oleg Cohen | Principal | **A S S U R E B R I D G E*
*Office: +1 617 564 0737 | Mobile: +1 617 455 7927 | Fax: +1 888 409
6995*
*Email: Oleg....@assurebridge.com <Oleg....@assurebridge.com> **|
www.assurebridge.com <http://www.assurebridge.com>*

Jan Lehnardt

unread,
May 6, 2016, 6:33:42 PM5/6/16
to us...@couchdb.apache.org
Oleg, thanks for the report! This was reported before, but only very
recently and we are still looking into it. Can you maybe add your setup
details to the ticket at:

https://issues.apache.org/jira/browse/COUCHDB-3009?

Thank you!

Best
Jan
--
--
Professional Support for Apache CouchDB:
https://neighbourhood.ie/couchdb-support/

Robert Samuel Newson

unread,
May 8, 2016, 6:48:53 PM5/8/16
to user
would be useful to get the response to GET :5984/_membership from both nodes during this period.

B.
Reply all
Reply to author
Forward
0 new messages