delete_misses increases
I am running a cluster of 3 couchbase nodes.
A client side moxi is configured as below (other params default)
----------------------
downstream_max=65535,
downstream_conn_max=4
---------------------
freeswitch is client and requests key/values to moxi.
I am using sipp to make bulk calling. After some time, sipp retransmits calls.
and when looked into couchbase stats using cbstats command, I got below,
---------------
delete_misses = 124587
---------------
as this counter increases, retransmission of the calls also increases.
Wt do I do ? Do I miss any configuration ?
thnx for the reply mikew
I never delete a key in couchbase. In my application once I set keys/values, I use them to read only.
The same, when I use memcached server, I dont get any delete misses.
Can you post all of your stats and describe your workload? I know you said you do only gets and sets but how many per second and what is the ratio of each. I only ask because if the stats don't show anything I will try to reproduce this problem to help answer your question. Stats can be gotten with the command below.
/opt/couchbase/bin/cbstats :11210 all
or
/opt/couchbase/bin/cbstats :11210 all -b
A delete miss means that the item your trying to delete doesn't exist. This means you either already deleted it or you never set a value for that key. If your client is retrying upon a delete miss then it sounds like there may be an issue in your client logic.