Speculative retry

[Tommy Stendahl]

Hi,

We have some concern regarding speculative retry in the driver.

If we have a slow request and trigger a speculative retry and the first
response we get back is a failure (maybe we didn't reach CL or something
else), this will trigger the driver to respond back to the client with
this failure. But maybe we get a success full response from the other
request, then the driver have responded with a failure even though the
request was successful. Is this good?

In some use cases it might make sense to have a retry policy that
ignores the first response if it is a failure and waits for the next one
if there is more then one outstanding copy of the request.

Any thoughts on this?

Thanks
Tommy

[Alexandre Dutra]

Hi Tommy,

I suppose you are familiar with speculative executions and retries. So if I follow you correctly, you are suggesting that the driver should keep waiting for more executions to complete when it believes that the error is non-permanent, is that right? 

That would blur the boundaries between these notions, but I see your point. We could indeed re-use some of the logic currently living in retry policies to decide whether or not to stop the current round of speculative executions based on the error received.

I would suggest that you open a Jira ticket with your proposal so that we can investigate further.

Thanks,

Alexandre

--
You received this message because you are subscribed to the Google Groups "DataStax Java Driver for Apache Cassandra User Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to java-driver-us...@lists.datastax.com.

--

Alexandre Dutra

Driver & Tools Engineer @ DataStax

[Tommy Stendahl]

Hi Alexandre,

Yes, I think you understand me correctly, that’s exactly what I would like do.

I have created a jira, JAVA-1440.

Thanks, Tommy