Hi, I'm going to migrate my flink cluster from 1.4.0 to 1.5.3, and I have been trying to map config file to the latest version. I used to use these three configuration. Are they still needed in FLIP-6 mode? Moreover, is any akka config still needed in FLIP-6 mode? Since I had a impression that FLIP-6 tried to get rid of akka and use its own rpc interface. Please correct me if I misunderstood. Thanks. akka.watch.heartbeat.interval akka.watch.heartbeat.pause taskmanager.exit-on-fatal- Best Regards, Tony Wei |
Hi Tony, You are right that with FLIP-6 Akka is abstracted away. If you want custom heartbeat settings, you can configure the options below [1]: heatbeat.interval heartbeat.timeout The config option taskmanager.exit-on-fatal-akka-error is also not relevant anymore. I closest I can think of is taskmanager.registration.timeout [2]. Best, Gary [1] https://ci.apache.org/projects/flink/flink-docs-release-1.5/ops/config.html#heartbeat-manager [2] https://ci.apache.org/projects/flink/flink-docs-release-1.5/ops/config.html#taskmanager-registration-timeout On Mon, Sep 10, 2018 at 4:24 AM, Tony Wei <[hidden email]> wrote:
|
I should add that in FLIP-6 mode we are not relying on Akka's DeathWatch but because Flink's RPC framework uses Akka, you are still able to configure the other Akka config options [1]. [1] https://ci.apache.org/projects/flink/flink-docs-release-1.5/ops/config.html#distributed-coordination-via-akka On Mon, Sep 10, 2018 at 7:38 AM, Gary Yao <[hidden email]> wrote:
|
Hi Gray, Thanks for your useful information! Here I wonder if the following configs still valid on FLIP-6 mode. 1. akka.transport.heartbeat.interval 2. akka.transport.heartbeat.pause It seems they are different from HeartbeatServices and possibly still valid. Best, tison. Gary Yao <[hidden email]> 于2018年9月10日周一 下午1:50写道:
|
Hi Tison, These can be still be set but judging from the documentation [1], they have never been very relevant in Flink: Heartbeat interval for Akka's transport failure detector. Since Flink uses TCP, the detector is not necessary. Therefore, the detector is disabled by setting the interval to a very high value. In case you should need the transport failure detector, set the interval to some reasonable value. The interval value requires a time-unit specifier (ms/s/min/h/d). If you have a good reason to use the transport failure detector, I would be interested in it. Best, Gary [1] https://ci.apache.org/projects/flink/flink-docs-release-1.5/ops/config.html#akka-transport-heartbeat-interval On Mon, Sep 10, 2018 at 8:20 AM, 陈梓立 <[hidden email]> wrote:
|
Free forum by Nabble | Edit this page |