Hi Jake,
Thanks for the reply. It appears that my task fails too many times within
the yarn.container.retry.window.ms
​
window. Increasing yarn.container.retry.count​seems to solve the issue for
the time being.
It still remains unknown why my tasks are failing so often. There is no
error messages
Hi all,
I couldn't find in the documentation (
https://samza.apache.org/learn/documentation/latest/jobs/configuration-table.html)
where on the disk Samza stores the local RocksDB key-value store. Can
anyone tell me where it is and if it can be configured?
Thanks,
Jack
> On June 9, 2016, 7:15 p.m., Chris Pettitt wrote:
> > Thoughts below. A more general thought: the builders don't seem to be
> > providing a lot of value and they're constraining you to a bad API (string
> > based, untyped). I'd strongly suggest that we use typed builders and map
> > this back
Hey Jack,
The port collisions are fixed in
https://issues.apache.org/jira/browse/SAMZA-932. However, they usually
don't occur frequently enough to cause the job to fail. Usually only a
couple containers fail with this exception but the job restarts them and
eventually runs smoothly.
Your GC log l
---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/48080/
---
(Updated June 10, 2016, 7:39 p.m.)
Review request for samza.
Changes
---
Hi all,
I am having difficulty determining the reason my Samza task is failing. It
generally failed within 10 minutes of start. When I examine the YARN log I
see the following exception on some but not all containers:
java.rmi.server.ExportException: Port already in use: 40029; nested
exception i
