[
https://issues.apache.org/jira/browse/CASSANDRA-12791?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15621787#comment-15621787
]
Sylvain Lebresne commented on CASSANDRA-12791:
----------------------------------------------
bq. The intent is to help operators work out if messages are dropped because of
clock skew.
bq. we need to also check DD.hasCrossNodeTimeout(), a message originating cross
node is not sufficient.
I don't I agree tbh. Adding the {{DD.hasCrossNodeTimeout()}} check is losing
information and creating a somewhat confusing metric, but I disagree it's
really adding value. To quote Brandon on the original ticket, knowing if
messages are dropped of clock skew "is easily derived from the yaml". Namely,
if you do see a lot of cross-node dropped message but no local/internal ones,
then it's a fair sign this may be due to clock skew and you can then simply
check if {{DD.hasCrossNodeTimeout()}} is set or not to confirm.
So adding the {{DD.hasCrossNodeTimeout()}} check does not really add any
information that you can't easily infer otherwise, but adding it does mean that
when the option is {{false}} (the default as it happens), then the cross-node
metric will never-ever get incremented. And I can't shake the feeling that it's
going to be confusing for most users.I mean, they see we have 2 different
metrics, but only seeing lhe "local" one ever get incremented might make them
think only locally delivered message are dropped for some weird reason.
Anyway, I don't care tremendously about it (I was mostly bugged by the broken
logic in {{MessageIn}} after all) but I do think it's strictly better *without*
the check to {{DD.hasCrossNodeTimeout()}} in {{MS.incrementDroppedMessage()}}.
I'm good with the rest of the changes though.
> MessageIn logic to determine if the message is cross-node is wrong
> ------------------------------------------------------------------
>
> Key: CASSANDRA-12791
> URL: https://issues.apache.org/jira/browse/CASSANDRA-12791
> Project: Cassandra
> Issue Type: Bug
> Reporter: Sylvain Lebresne
> Assignee: Sylvain Lebresne
> Priority: Minor
>
> {{MessageIn}} has the following code to read the 'creation time' of the
> message on the receiving side:
> {noformat}
> public static ConstructionTime readTimestamp(InetAddress from, DataInputPlus
> input, long timestamp) throws IOException
> {
> // make sure to readInt, even if cross_node_to is not enabled
> int partial = input.readInt();
> long crossNodeTimestamp = (timestamp & 0xFFFFFFFF00000000L) | (((partial
> & 0xFFFFFFFFL) << 2) >> 2);
> if (timestamp > crossNodeTimestamp)
> {
> MessagingService.instance().metrics.addTimeTaken(from, timestamp -
> crossNodeTimestamp);
> }
> if(DatabaseDescriptor.hasCrossNodeTimeout())
> {
> return new ConstructionTime(crossNodeTimestamp, timestamp !=
> crossNodeTimestamp);
> }
> else
> {
> return new ConstructionTime();
> }
> }
> {noformat}
> where {{timestamp}} is really the local time on the receiving node when
> calling that method.
> The incorrect part, I believe, is the {{timestamp != crossNodeTimestamp}}
> used to set the {{isCrossNode}} field of {{ConstructionTime}}. A first
> problem is that this will basically always be {{true}}: for it to be
> {{false}}, we'd need the low-bytes of the timestamp taken on the sending node
> to coincide exactly with the ones taken on the receiving side, which is
> _very_ unlikely. It is also a relatively meaningless test: having that test
> be {{false}} basically means the lack of clock sync between the 2 nodes is
> exactly the time the 2 calls to {{System.currentTimeMillis()}} (on sender and
> receiver), which is definitively not what we care about.
> What the result of this test is used for is to determine if the message was
> crossNode or local. It's used to increment different metrics (we separate
> metric local versus crossNode dropped messages) in {{MessagingService}} for
> instance. And that's where this is kind of a bug: not only the {{timestamp !=
> crossNodeTimestamp}}, but if {{DatabaseDescriptor.hasCrossNodeTimeout()}}, we
> *always* have this {{isCrossNode}} false, which means we'll never increment
> the "cross-node dropped messages" metric, which is imo unexpected.
> That is, it is true that if {{DatabaseDescriptor.hasCrossNodeTimeout() ==
> false}}, then we end using the receiver side timestamp to timeout messages,
> and so you end up only dropping messages that timeout locally. And _in that
> sense_, always incrementing the "locally" dropped messages metric is not
> completely illogical. But I doubt most users are aware of those pretty
> specific nuance when looking at the related metrics, and I'm relatively sure
> users expect a metrics named {{droppedCrossNodeTimeout}} to actually count
> cross-node messages by default (keep in mind that
> {{DatabaseDescriptor.hasCrossNodeTimeout()}} is actually false by default).
> Anyway, to sum it up I suggest that the following change should be done:
> # the {{timestamp != crossNodeTimestamp}} test is definitively not what we
> want. We should at a minimum just replace it to {{true}} as that's basically
> what it ends up being except for very rare and arguably random cases.
> # given how the {{ConstructionTime.isCrossNode}} is used, I suggest that we
> really want it to mean if the message has shipped cross-node, not just be a
> synonymous of {{DatabaseDescriptor.hasCrossNodeTimeout()}}. It should be
> whether the message shipped cross-node, i.e. whether {{from ==
> BroadcastAdress()}} or not.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
