Multiple servers Multiple Issues.

[jdarinthomas]

I have a multiple server setup.
I have two issues. I don't think they are related, but I'm posting them together just in case they are.

Server 1 is on it's own Ubuntu 20.04 physical machine.
Hardware is:
CPU (6 core Xeon), ram (24Gb) OS drive is SATA SSD and data drive is WD 2tb purple.
It has a dedicated nic for inter-server data network, a dedicated nic for the camera network, and a dedicated nic for the user network.
This server hosts the db and data drive. In addition to the db and data drive, this server has 2 active lower resolution cameras and 4 decommissioned cameras (archived videos that I want associated with the retired cameras).
Load in the server is light:
CPU less than 20%
Real memory less than 15%
/dev/shm less than 15%

Servers 2, 3, and 4 are VM's running Ubuntu 22.04. Host is Ubuntu 22.04 running KVM (the Cloudmin flavor)
They share a dedicated nic for the inter-server data network, a dedicated nic for the camera network and a dedicated nic for the user network.
Loads are medium ish.
Server 2
CPU less than 30%
Real memory less than 40%
/dev/shm less than 50%
Server 3
CPU less than 75%
Real memory less than 50%
/dev/shm less than 60%
Server 4
CPU less than 80%
Real memory less than 50%
/dev/shm less than 70%

First issue:
3 cameras that used to be assigned to server 1, but now are assigned to other servers. This error is generated the "Can't open memory map file /dev/shm/zm.mmap.XX: No such file or directory" in the log with server 1 listed in the server column. In the database table "Monitors" they all have the correct new server ID. If i reassign the cameras back to server 1 the error goes away.

Second Issue:
I have events where the event name is "New Event" instead of "Event-eventID", these events have no end time and are unplayable.

Thanks for any assistance

Darin

[dougmccrary]

Second Issue:
I have events where the event name is "New Event" instead of "Event-eventID", these events have no end time and are unplayable.

That indicates in progress, or either abnormal exit or crash, in my experience. Unless there are more than a couple of them per monitor, running zmaudit.pl should clean them up. If there are a lot, then worth investigating - it's not clear where the zm db is - unless space is tight, it should be on the ssd.

[jdarinthomas]

So, how do I chase these crashes, they occur often and are not recordings in progress. By often, I mean greater than 60% of events are recordings that crashed.

Could it be related to the "type" of processor presented to the VM?

The zm db is on the SSD, the only thing on the data drive is /var/cache/zoneminder/*

[dougmccrary]

You start by looking at the logs. Concentrate on one monitor using the search filter.
Might have to turn on debug if that doesn't help.