i am doing bulk upserts to a sharded cluster with bulks of 1000. Everything
seems to work fine, except that in the middle of the script it suddenly
pauses - sometimes just minutes, sometimes half an hour - and then proceeds
Could you provide more information:
- What do the logs show during this pause period? (i.e. mongod or mongos
logs), and what is the exact command you are executing?
- Please provide the output of sh.status() (to see your shard key, and
- What is your MongoDB server version, and client driver version, if
- What is your configured storage engine and its options
- What is your O/S and the filesystem that contains the /data/db
- What is your deployment topology (i.e. how many shards, do the shards
consist of a replica set each, how many config servers, how many mongos,
Also, can you share a code snippet that performs the bulk operation, along
with any write concern for the operation?
I tried profiling for long running methods but nothing showed up. I also
looked at the server stats and it seems that the server is doing some cpu
work, but no io.
A couple of questions:
- How and on which server did you turn on the profiler?
- On which server and using what command did you run the server stats?
Could you show us some example output?
You received this message because you are subscribed to the Google Groups "mongodb-user"