Hi Jonathan, That shouldn't happen. The API returns the answer from the Overseer node (via ZK) and should return immediately. The API will timeout after 180s if somehow it cannot get a response from Overseer. I don't see why it would timeout. What's the read timeout on your monitoring system?
About your shards getting marked as down, do you send commits explicitly? If yes, you may be running into SOLR-6530 which is fixed in 4.10.2. On Mon, Dec 8, 2014 at 9:45 PM, Hutchins, Jonathan <jhutch...@webmd.net> wrote: > We are currently running Solr 4.10.0 in production. We have run into an > issue where we cannot have our monitoring system hit the CLUSTERSTATUS api > command every five minutes (or even as long as every hour) without getting > a significant number of timeouts on the command. Does this command return > a timeout message after a period of time (and we aren’t waiting long > enough) or will it hang indefinitely? This is a major problem for us > because we are finding that shards will randomly get marked down, turn red > in the web console, but there is no indication that they are down except if > we run that command. Please help! > > Thanks! > > - Jonathan > -- Regards, Shalin Shekhar Mangar.
