Query statistics

[doublex]

Is there a button to query nebula-python if a query was exhausting?
Nebula is never slow - but I have the impression some queries are CPU expensive.
I would like to pinpoint those.

[wey-gu]

Is there a button to query nebula-python if a query was exhausting? Nebula is never slow - but I have the impression some queries are CPU expensive. I would like to pinpoint those.

Would you mind describing more on your picture of this feature? Better with some reference that other lib provided similar functions?

[doublex]

Yes, sorry.
Some query is causing a lot of CPU load on the Nebula server.
Is there a profiler tool to find out which query it is?
My first thought was a summary-feature like:

result = session.execute( statement )
slow_query = result.get_stats()

[wey-gu]

You could go with SHOW QUERIES, for now, we could only see like how long they were executed.

But normally long run queries are those we need to care about, right?

[doublex]

@wey-gu
Thanks for everything!!

[wey-gu]

np, @doublex will SHOW QUERIES and KILL QUERY help?

[doublex]

You put me on the right track:
https://docs.nebula-graph.io/3.4.0/3.ngql-guide/17.query-tuning-statements/1.explain-and-profile/
Best wishes!

[wey-gu]

Wow good to know it helped in some way.

Then I would recommend my blog post about this: https://www.siwei.io/ngql-execution-plan/ while it's Chinese only for now, if you can't read Chinese, I'll post the English version soon.(sorry about this)

[doublex]

@wey-gu
Impressive! Thanks for everything!