I have a script that connects to my local mongos and do some processing. However, during the process, I encountered many cursor timeout exceptions, which (I suspected) left many dead connections on my server.
When I run mongostat
on my host, I see 450+ connections:
insert query update delete getmore command vsize res faults netIn netOut conn repl time
0 0 0 0 0 1 977m 195m 0 62b 726b 458 RTR 02:11:07
0 0 0 0 0 1 977m 195m 0 62b 726b 458 RTR 02:11:08
0 0 0 0 0 1 977m 195m 0 62b 726b 458 RTR 02:11:09
0 0 0 0 0 1 977m 195m 0 62b 726b 458 RTR 02:11:10
How can I make sure that these are idle resources that I should free up? If they are, how can I free them?
Best Answer
This should be taken care programmatically. Programmers have to close the connection whenever they are done with dB access else memory leaks could happen.
I'm not sure how these processes can be killed as an admin.