Incorrect failure handler exceptions

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Incorrect failure handler exceptions

Ilya Kasnacheev
Hello!

https://issues.apache.org/jira/browse/IGNITE-13665

The notorious problem with failure handler's thread dump, which nobody has
bothered to investigate, until now. "IgnitionEx$IgniteNamedInstance$2.apply"

If anybody knows how to fix it, please suggest.

Regards,
--
Ilya Kasnacheev
Reply | Threaded
Open this post in threaded view
|

Re: Incorrect failure handler exceptions

Ivan Daschinsky
There is not a big problem to fix it.

1. Blocked thread is detected by another worker. Currently we take dump
before printing it to log with zero depth. This can be easily fixed.

2. We should propagate BlockedThreadExceptiob to failureCtx with original
stacktrace of blocked thread. StackFrames, obtained in previous step, can
be used to construct stacktrace of BlockedThreadException.


вт, 3 нояб. 2020 г., 19:15 Ilya Kasnacheev <[hidden email]>:

> Hello!
>
> https://issues.apache.org/jira/browse/IGNITE-13665
>
> The notorious problem with failure handler's thread dump, which nobody has
> bothered to investigate, until now.
> "IgnitionEx$IgniteNamedInstance$2.apply"
>
> If anybody knows how to fix it, please suggest.
>
> Regards,
> --
> Ilya Kasnacheev
>
Reply | Threaded
Open this post in threaded view
|

Re: Incorrect failure handler exceptions

Ilya Kasnacheev
Hello!

The fix is now merged.

Regards,
--
Ilya Kasnacheev


вт, 3 нояб. 2020 г. в 22:51, Ivan Daschinsky <[hidden email]>:

> There is not a big problem to fix it.
>
> 1. Blocked thread is detected by another worker. Currently we take dump
> before printing it to log with zero depth. This can be easily fixed.
>
> 2. We should propagate BlockedThreadExceptiob to failureCtx with original
> stacktrace of blocked thread. StackFrames, obtained in previous step, can
> be used to construct stacktrace of BlockedThreadException.
>
>
> вт, 3 нояб. 2020 г., 19:15 Ilya Kasnacheev <[hidden email]>:
>
> > Hello!
> >
> > https://issues.apache.org/jira/browse/IGNITE-13665
> >
> > The notorious problem with failure handler's thread dump, which nobody
> has
> > bothered to investigate, until now.
> > "IgnitionEx$IgniteNamedInstance$2.apply"
> >
> > If anybody knows how to fix it, please suggest.
> >
> > Regards,
> > --
> > Ilya Kasnacheev
> >
>
Reply | Threaded
Open this post in threaded view
|

Re: Incorrect failure handler exceptions

Ivan Daschinsky
Ilya, thanks for your effort.

ср, 11 нояб. 2020 г. в 14:28, Ilya Kasnacheev <[hidden email]>:

> Hello!
>
> The fix is now merged.
>
> Regards,
> --
> Ilya Kasnacheev
>
>
> вт, 3 нояб. 2020 г. в 22:51, Ivan Daschinsky <[hidden email]>:
>
> > There is not a big problem to fix it.
> >
> > 1. Blocked thread is detected by another worker. Currently we take dump
> > before printing it to log with zero depth. This can be easily fixed.
> >
> > 2. We should propagate BlockedThreadExceptiob to failureCtx with original
> > stacktrace of blocked thread. StackFrames, obtained in previous step, can
> > be used to construct stacktrace of BlockedThreadException.
> >
> >
> > вт, 3 нояб. 2020 г., 19:15 Ilya Kasnacheev <[hidden email]>:
> >
> > > Hello!
> > >
> > > https://issues.apache.org/jira/browse/IGNITE-13665
> > >
> > > The notorious problem with failure handler's thread dump, which nobody
> > has
> > > bothered to investigate, until now.
> > > "IgnitionEx$IgniteNamedInstance$2.apply"
> > >
> > > If anybody knows how to fix it, please suggest.
> > >
> > > Regards,
> > > --
> > > Ilya Kasnacheev
> > >
> >
>


--
Sincerely yours, Ivan Daschinskiy