On Tue, Sep 26, 2017 at 6:26 PM, Atul Gosain <agos...@luminanetworks.com> wrote:
> Controller devs > > We have been facing issues while trying scaled scenarios on bgp and > netconf projects when the setup was clustered. We had tried some settings > to improve the replication performance like using tell-based mode enabled > and then tried to tinker with few other factory/akka.conf settings like > akka.cluster.failure.detector.acceptable-heartbeat-pause. These measures > improved performance a bit since the tolerance for heartbeat detection > failure was increased. > > We also enabled the artery mode to and did the corresponding setting > changes in akka.conf to enable it. After some other setting changes, we > could avoid the OOM issues that we faced initially. But inspite of these, > we were not able to successfully run the cluster on artery mode. > > We may be able to troubleshoot it further if we are able to find some more > detailed logs from akka. > > Are there any > 1. Log settings we can enable to get the debug/trace level logs from akka ? > You can set akka.loglevel=DEBUG in the akka.conf and also http://doc.akka.io/docs/akka/2.4.0/scala/logging.html#Auxiliary_remote_logging_options > 2. Any other akka settings which we can tune to choose an acceptable > tradeoff ? > > We have been using master for these tests to bring in the patches to > improve cluster performance (mainly from TomP and Robert). > > -- > > Thanks > Atul > > _______________________________________________ > controller-dev mailing list > controller-dev@lists.opendaylight.org > https://lists.opendaylight.org/mailman/listinfo/controller-dev > >
_______________________________________________ controller-dev mailing list controller-dev@lists.opendaylight.org https://lists.opendaylight.org/mailman/listinfo/controller-dev
