[
https://issues.apache.org/jira/browse/CASSANDRA-15642?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17077619#comment-17077619
]
Benedict Elliott Smith commented on CASSANDRA-15642:
----------------------------------------------------
{quote}I propose waiting for {{blockFor}} (with a {{CountDownLatch}} for
example in {{ReadCallback}}, as detailed on CASSANDRA-15543), until timeout, if
it happens.
{quote}
What about on success? Should we also wait to see if everyone responds, or if
a node is having a problem to report? Why do we only want to hear about these
problems when the operation fails?
{quote}I would argue that the user wouldn't be able to act optimally anyway
with only partial information. In the case of "failures + timeouts" then there
would be more value in them knowing that there is also timeouts with some
nodes, instead of just the failure.
{quote}
Is that an argument, or a belief?
I'm personally not sure how I would utilise this extra information. I have no
idea from it if the timeouts are caused by something transient or serious, nor
what infrastructure might have caused it. It could be the coordinator itself,
the peer, a link between them. More importantly, writing software to
intelligently use this information automatically is... ambitious. Mostly this
kind of information is IME useful for human diagnosis, but for this I would be
interested in logs and metrics, and I would be interested in time periods and
many operations, and we have this kind of aggregate data elsewhere.
However the cost of delaying an answer to the client is potentially significant.
If you were to try to introduce this behaviour, it would have to be optional
IMO. This would be a huge semantic change to force upon our users
unconditionally.
> Inconsistent failure messages on distributed queries
> ----------------------------------------------------
>
> Key: CASSANDRA-15642
> URL: https://issues.apache.org/jira/browse/CASSANDRA-15642
> Project: Cassandra
> Issue Type: Improvement
> Components: Consistency/Coordination
> Reporter: Kevin Gallardo
> Priority: Normal
>
> As a follow up to some exploration I have done for CASSANDRA-15543, I
> realized the following behavior in both {{ReadCallback}} and
> {{AbstractWriteHandler}}:
> - await for responses
> - when all required number of responses have come back: unblock the wait
> - when a single failure happens: unblock the wait
> - when unblocked, look to see if the counter of failures is > 1 and if so
> return an error message based on the {{failures}} map that's been filled
> Error messages that can result from this behavior can be a ReadTimeout, a
> ReadFailure, a WriteTimeout or a WriteFailure.
> In case of a Write/ReadFailure, the user will get back an error looking like
> the following:
> "Failure: Received X responses, and Y failures"
> (if this behavior I describe is incorrect, please correct me)
> This causes a usability problem. Since the handler will fail and throw an
> exception as soon as 1 failure happens, the error message that is returned to
> the user may not be accurate.
> (note: I am not entirely sure of the behavior in case of timeouts for now)
> For example, say a request at CL = QUORUM = 3, a failed request may complete
> first, then a successful one completes, and another fails. If the exception
> is thrown fast enough, the error message could say
> "Failure: Received 0 response, and 1 failure at CL = 3"
> Which:
> 1. doesn't make a lot of sense because the CL doesn't match the number of
> results in the message, so you end up thinking "what happened with the rest
> of the required CL?"
> 2. the information is incorrect. We did receive a successful response, only
> it came after the initial failure.
> From that logic, I think it is safe to assume that the information returned
> in the error message cannot be trusted in case of a failure. Only information
> users should extract out of it is that at least 1 node has failed.
> For a big improvement in usability, the {{ReadCallback}} and
> {{AbstractWriteResponseHandler}} could instead wait for all responses to come
> back before unblocking the wait, or let it timeout. This is way, the users
> will be able to have some trust around the information returned to them.
> Additionally, an error that happens first prevents a timeout to happen
> because it fails immediately, and so potentially it hides problems with other
> replicas. If we were to wait for all responses, we might get a timeout, in
> that case we'd also be able to tell wether failures have happened *before*
> that timeout, and have a more complete diagnostic where you can't detect both
> errors at the same time.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
