Please re-open this issue [CORE2645]

[firebird-automations]

Submitted by: Hannes Lowette (hanneslowette)

Jira_subtask_inward CORE2462

2.5 beta 2 still has this issue

[firebird-automations]

Commented by: @dyemanov

Please take a look at DNET275, wait for a fixed build (or compile the binaries from SVN yourself) and report back whether the issue still exists with a new version of the .NET provider.

[firebird-automations]

Commented by: Hannes Lowette (hanneslowette)

We still have this using version 2.5.2 of the .NET provider.

Symptoms:

- Update some record(s)
- Commit

--> engine goes to 100% CPU, without even stopping the process or disconnecting the connection.
--> client application becomes unresponsive.

[firebird-automations]

Commented by: Hannes Lowette (hanneslowette)

Maybe some useful info:

If we kill the engine and restart it, the updates/inserts seem to be committed.

[firebird-automations]

Commented by: @dyemanov

I'm getting confused. The original issue was about the fast growing log file, now you talk about the client application hanging and the engine using 100% of CPU. Any hint what should I answer you? :-)

BTW, try the latest (today's) FB 2.5 snapshot. I'd expect it to have the firebird.log issue resolved finally.

[firebird-automations]

Commented by: Hannes Lowette (hanneslowette)

The server goes to 100% CPU and starts writing a log file like crazy (exactly as detailed in the original issue). I'm guessing it's writing to the log file non stop. The engine becomes otherwise non-responsive (which is probably because the logfile priority is higher than the rest of the FB process), so it results in hanging clients.

This is by all means thesame issue.

I'll give the snapshot build a go and I'll let you know if it helped.

Thanks for looking into this.