[
https://issues.apache.org/jira/browse/IGNITE-25161?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Roman Puchkovskiy updated IGNITE-25161:
---------------------------------------
Description:
If an exception happens during node start/join, we don't log it ourselves in
IgniteImpl/IgniteServerImpl, instead we return it via start/join future to the
code that initiated the start/join. But this code might just discard the
exception or log it in an ephemeral way (print to stdout instead of writing to
log file, for example). This is what IgniteRunner does.
To make sure we don't lose the startup failure root cause, we should also log
it internally, before returning to the caller.
> Improve exception logging during node start
> -------------------------------------------
>
> Key: IGNITE-25161
> URL: https://issues.apache.org/jira/browse/IGNITE-25161
> Project: Ignite
> Issue Type: Improvement
> Reporter: Roman Puchkovskiy
> Assignee: Roman Puchkovskiy
> Priority: Major
> Labels: ignite-3
> Time Spent: 10m
> Remaining Estimate: 0h
>
> If an exception happens during node start/join, we don't log it ourselves in
> IgniteImpl/IgniteServerImpl, instead we return it via start/join future to
> the code that initiated the start/join. But this code might just discard the
> exception or log it in an ephemeral way (print to stdout instead of writing
> to log file, for example). This is what IgniteRunner does.
> To make sure we don't lose the startup failure root cause, we should also log
> it internally, before returning to the caller.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
