When the query fails, can you run the same query directly to Pinot? Let's first find out if isue is Pinot side or Presto side.

@Mayank how should I submit 4 queries simentenoulsly with Pinot UI.because this issue I am getting when I was trying to run all the queries same time. If I am running queries seperately then it's working fine.

Same queries running separately works fine. or those are different queries?

Same queries working fine separately

What's the resources (CPU/MEM/HEAP) on pinot vms?

I am only facing the issue when I am refreshing the entire dashboard where I used all these four queries.

One possibility is that the queries are expensive when run in parallel and your Pinot VMs do not have enough resource to answer the query in parallel.

Hmm any way to check resource allocation in pinot. Queries are also not expensive just taking sum and doing group by on one column. In 2nd query taking count on same group by column. And in 3rd query calculating percentage on same group by column. And I have only 1000 records in table.

I am asking about VM size heap size etc. If you are using latest master there's an appConfig endpoint in Swagger.

I m on swagger UI where I need to check that

Is this latest master, or 0.7.1?

appConfigs is not there. I have 0.7 1

yeah that code does not have it

Ok let me check

And also for the individual successful queries, let's see the numDocs/numEntries etc. Then we will know how much work each query is doing and how much resource is there on VM.

Ok @Mayank

I feel it’s timeout on pinot side?

Yeah, that's what I am suspecting too.

you can view it from the right link button when you enter into each chart

Only when these queries fired in parallel, that's when this happens.

then maybe try to increase pinot server timeout