Couchbase 6 Index scan timeout

SQL++
, ,
1

Hi everybody,
I’m working on a Couchbase 6 cluster with 5 nodes.
I have a Bucket with 10M documents and 16GB ram quota.
I perform an aggregation query using python sdk over an index.
Usually I am able to extracts documents: Get 638605 documents in 90.60135912895203 seconds
Sometimes I get:

couchbase.exceptions.CouchbaseException: <RC=0xCD[LCB_ERR_INTERNAL_SERVER_FAILURE (205)], HTTP Request failed Index scan timed out - cause: Index scan timed out

There is no way for me to improve index or query so I would like to find a way to increase the default index timeout on Couchbase version 6 but I don’t know how.

Any suggestions? Thanks in advance

2

@amarino , it would be good to check the memory quota allocated to the index service. If the index resident percent in memory is less, it will need more disk IO and slow down the scan.

You can increase the scan timeout using below settings change:

curl -X POST -u user:pass http://index_node_ip:9102/settings --data ‘{“indexer.settings.scan_timeout”: <timeout_in_milliseconds>}’

1 Like
3

If EE see if you can use index aggreation.

https://index-advisor.couchbase.com/indexadvisor/#1

1 Like