Getting error on query: All indexer replica is down or unavailable or unable to process request - cause: queryport.client.noHost

Hello!

I have a shared index on a cluster created with query:

CREATE INDEX `idx_errors` 
ON `data`((distinct (array `err` for `err` in (`extract`.`errors`) end))) 
PARTITION BY hash((meta().`id`)) WITH { "num_partition":8 }

And I am trying to query data the with following query:

SELECT count(*) as count_errors 
FROM data AS doc 
WHERE ANY e IN doc.extract.errors SATISFIES e LIKE "%host" END;

And getting error:

{
    "code": 5000,
    "msg": "All indexer replica is down or unavailable or unable to process request - cause: queryport.client.noHost"
}

I haven’t found any explanation of this error in the internet. Is there any additional info about how to fix this issue or what to do further?

@icamys

Can you kindly let us know the couchbase-server version you are using. Also, what is your cluster status. Is any of the indexer node on which the index “idx_errors” resides is inactive/failed over?

@varun.velamuri

Hello! Thank you for your response.

I am testing Enterprise Edition 6.0.1 build 2037. All servers respond, all indexers are in green status and healthy as far as I can see in web GUI.

Maybe there’s any other info that I can provide?

@icamys

Can you kindly share the indexer logs from indexer nodes where the index idx_errors resides and query logs so that we can investigate the issue further. The logs can be found in /opt/couchbase/var/lib/couchbase/logs directory on a linux distribution.

Thanks,
Varun

Hi, I’m DevOps from @icamys team, in the next message attached indexer and query logs with reference to idx_errors index.

[UPD: I can’t upload logs in reason of: ‘Sorry, new users can not upload attachments.’]

@rambling

Sorry for the delayed response. I missed your notification.

You can upload the logs on a google drive and share it to me on varun[dot]velamuri[at]couchbase[dot]com

Apologies for the delay again.

Thanks,
Varun

@varun.velamuri
Done. Please check mail.

@rambling,

Thanks for sharing the logs. From the logs, I see that there is no index for bucket data (rather I see an index for bucket scraped_data). Am I missing anything? Also, can you share the logs when the issue is happening?

Thanks,
Varun

@varun.velamuri
Data bucket is just for example. For real we work with idx_errors index on scraped_data bucket. I’ve resented extended logs from needed period.
Please check mail.

Hi @rambling,

The issue is a known issue in 6.0.1 version (See https://issues.couchbase.com/browse/MB-34003). It happens when ns_server changes the address of indexer from 127.0.0.1 to 144.76.78.11.

As a long term solution, I recommend you to upgrade to 6.0.2 where this issue is fixed. As a temporary solution, you can terminate the cbq-engine/indexer process on the node 144.76.78.11.

Thanks,
Varun