How can i schedule savepoint triggering from flink...
# troubleshooting
b
How can i schedule savepoint triggering from flink ?
p
How are you deploying Flink?
b
I am using K8
i can't find the config to scheudle savepoint
p
operator or pure k8s?
b
pure
image.png
my savepoints always none.
p
You'd have to have some sort of k8s job to run
Copy code
$ bin/flink savepoint :jobId [:targetDirectory]
on your JM pod. Docs
b
how about RETAIN_ON_CANCELLATION mode?
p
Or via REST API
b
if job manager down, will it save ?
p
how about RETAIN_ON_CANCELLATION mode?
This is about keeping checkpoints if you cancel the job
if job manager down, will it save ?
Did not get it
b
if JM pod down, will it restore the last checkpoint if I up back ?
p
If you have HA set in place, I'd say so.
b
i will give a try. But when i enable HA (> 1 Job manager on the run) their query response rate will slow down due to finding the job manager leader.
ok, i change to HA cluster. it seemed better control. now it restored
p
Nice!
b
hi @Pedro Mázala, do you have any idea how to connect to Paimon Slack ? We found flink with paimon is extrmely good performance for analytics.
p
I don't know about a Paimon slack. Checked their website but it links back to ASF slack
b
no problem. Thanks ! 🙂