Geek Answers Handbook

Heart-beat CouchDB Timeout

My CouchDB (production) server sometimes shuts down, without my desire. When checking the logs, I noticed that Erlang / CouchDB (one or the other) sends a heart rate, and this heart rate is controlled. When a timeout occurs, the server is brutally killed ( couchdb -k). This is almost always an undesirable behavior, specifically on a production server.

I ran into this problem , which states the same thing and indicates that this happens mainly during heavy server loads, which I also noticed.

I am looking for a solution to keep my server running, but for this I need to learn more about this behavior. Unfortunately, the wiki wiki and manual do not provide much information about this (only a parameter in the change API).

Can anyone provide more information about this?

+5
source share
2 answers

couchdb -k should kill the background process, and then the couch needs to be restarted. However, the restart function is disabled by default. To enable it, you must use the -r option with a value greater than 0. You can do this in the bin / couchdb script - change RESPAWN_TIMEOUT from 0 to, for example. 5.

+6
source

, - , erlang vm.

http://erlang.org/doc/man/erl.html http://erlang.org/doc/man/heart.html.

couch, , script "erl something -heart" grep .

, . ?

+1

More articles:


All Articles