Attention is currently required from: laforge, pespin.
1 comment:
Patchset:
I'm not sure it makes sense to have this kind of metric in regular use of the software
I am not sure either, but I don't think it would hurt a lot.
only for really specific adhoc debugging.
Well, we know that `gen_sctp:send` is blocking. And with the current architecture, it's possible that it blocks the whole process, be it `sctp_server` or `sctp_client` module. If this happens, we would at least **see** that it's because of `gen_sctp:send` taking too long.
Harald proposed a different architecture, in which a separate process is spawned for each S1AP PDU. This approach is more robust, because `gen_sctp:send` would only be blocking sending of a single PDU, letting the S1GW to process other PDUs in the meantime. Once this is implemented, we can even get rid of this histogram.
In any case, this commit is missing rationale ...
I will upload an updated version soon.
To view, visit change 39666. To unsubscribe, or for help writing mail filters, visit settings.