Suddenly writing to /var/cache/zoneminder/events

[Andyrh]

For many years all events have been written to Store1 which is /zm/store1. Now ZM is trying to write events to /var/cache/zoneminder/events. I found out because it filled the boot drive and crashed ZM.

What would make it not write to the configured location?
The disk is no more full than it has been for a long time and store1 is writable. ZM is hands-off once setup, so the only config changes recently was disabling a monitor. This started 3 days ago. Before that the server had been up for 3 weeks.

[iconnor]

The reason should be logged. Any failure to write to the configured location will cause it to try again at the default, with lots of error messages in the log.

In 1.37 it will even try any other storage areas that may be available.

[Andyrh]

Would that log as "Cannot write to event folder /zm/store1" in my case? I cannot find that in the log.
Right now my log is full of cannot write errors because I changed the permissions to default.
Should I delete default? I know is a past install I deleted default, but it did not go away.

Thank you for the help.

[iconnor]

default is hard coded in. You can delete it, but internally it still exists. Maybe in 1.37 I'll finish removing it, but that's not going to happen in 1.36.

The errors should look like Failed creating event dir at and will be at Warning level.

[Andyrh]

The hint on where to start helped. From the syslog:

Code: Select all

Sep 20 11:00:09 watcher fwupd[458986]: 16:00:09:0672 FuContext            Failed to load SMBIOS: invalid DMI data size, got 3895 bytes, expected 7767
Sep 21 00:18:00 watcher systemd[1]: zoneminder.service: Failed with result 'oom-kill'.
Sep 21 02:45:08 watcher fwupd[471809]: 07:45:08:0910 FuContext            Failed to load SMBIOS: invalid DMI data size, got 3895 bytes, expected 7767
Sep 21 09:04:11 watcher fwupd[476639]: 14:04:11:0845 FuContext            Failed to load SMBIOS: invalid DMI data size, got 3895 bytes, expected 7767
Sep 21 16:58:08 watcher kernel: [1798081.519645] EXT4-fs error (device sda1): mpage_map_and_submit_extent:2512: comm kworker/u24:3: Failed to mark inode 184156836 dirty
Sep 21 16:58:08 watcher systemd[1]: zoneminder.service: Failed with result 'oom-kill'.
Sep 21 16:58:27 watcher zmc_m7[485961]: WAR [zmc_m7] [Failed creating event dir at /zm/store1]

Looks like an ext4 error.

If someone happens to know please post what the error is. I am going digging.

Edit:
This appears to be the real problem. 2 attempts to fix and no luck.
Sep 24 17:02:03 watcher kernel: [ 25.026706] pcieport 0000:00:03.0: PCIe Bus Error: severity=Corrected, type=Data Link Layer, (Transmitter ID)

Thank you,