16.1.3: Warning when check updates

[framura]

Hi,

when i check for updates I get following warning:

Warning: stream_socket_client(): unable to connect to unix:///var/run/configd.socket (Connection refused) in /usr/local/opnsense/mvc/app/library/OPNsense/Core/Backend.php on line 93


What does it means?

I also submit this problem with crash reporter.

TIA

Alex

[smajor]

I've seen this as well.  Click Services then Diagnostics.

Make sure configd is running.  If not, start it.

I found mine off (or crashed) just this morning when I selected to check for updates.

[packet loss]

I haven't seen this issue yet although it appears to be an issue that needs to be addressed. Been running OPNsense straight for 3 days without a problem. Maybe your System->Log File or one of the other log files could provide some valuable information to see when configd is crashing.

[franco]

We've seen this happening when the backend daemon handles larger workloads like IDS event logs and rules getting killed by the system, most likely related to RAM restrictions. So we're already investigating. First step was to add the daemon to the GUI service controls so that it can be restarted from there.

What are your box specifications and OPNsense setup (IDS, proxy)?

[smajor]

Mine is: Intel(R) Core(TM)2 Duo CPU E8400 @ 3.00GHz, 2 CPUs: 1 package(s) x 2 core(s), 4GB RAM. I have about 23% of RAM in use with the new IDS in use and a few items enabled.

It isn't a regular thing. I'll have an uptime for awhile before noticing it, often a week or more from what I can tell.

[franco]

The crash reporter info is not sufficient in these cases, if this happens we need dmesg and system.log output to pin it down properly. Any help here is appreciated.

[smajor]

I'll see what I can do. I actually turned off IDS because it was crashing then recovering (mostly except for configd and sshd) every few hours so it was invisible unless you were watching. I'm now going on 18 hours with no issues, so you're suspicion of high workloads is correct.

[smajor]

I may bow out of troubleshoot this for a bit, I found after a recent power outage my system didn't come up cleanly and my recent "me too" might be related to that.  However, my first post predated that outage so I was seeing something. I have a question about what I found, but will post a new thread so as not to hijack this one.

[smajor]

Here's my requested logs, but I'm still convinced I may have something going on with hardware - possibly my drive given the fsck errors I've had.

This weekend will probably be a config back up and reinstall with a replacement drive.

[franco]

Thank you, I'll review this and report back.