No image on new monitors

[mconfortino]

Hi everybody and greetings from Patagonia. We have this site recording over a hundred cameras on v1.36.7. The server is a dual Xeon with a total of 40 cores and 128 GB of RAM paired with a 24 TB IBM Storwize storage. It runs Ubuntu Server 20.04. The system runs flawlessly, with PTZ on Axis and Bosch domes and capturing from Pelco, Sony and Vivotek cameras. We had to move a Hikvision camera physically and logically, so we moved the camera to it's new location and changed ip address. Now the monitor states it is working, at the prior fps rate, but there is no image. Deleted the monitor, created again and the same: no errors, running but no image. Created a new monitor with other camera and it is running with no image. Couldn't find anything in the logs. The other monitors are working as usual. Any idea? Thank you very much.

[dougmccrary]

How about if you clone an identical camera? Does the clone work? Then edit the source to what you want for the relocated one.

[mconfortino]

I tried that also. I can see the live view from the cameras in the browser. I have three more Hikavision cameras running and recording in zm. Same model, same connection string (except ip address, of course). Still no image. zmc is running for that monitor, log is zero bytes long.

[dougmccrary]

So a clone of a working camera didn't work even before you changed the ip for the moved one?

[mconfortino]

Yes. Working camera placed in new location, changed ip, status running but no image. Then took a working camera of the same make and model, cloned it, adjusted ip, and the new monitor (clone of) running without image. Then a new camera of another make, set up as new monitor with the original ip of the moved camera, same result. As if I had hit a limit or something. Existing 107 cameras continue to work in record mode without any problem. I am out of ideas on this.

[dougmccrary]

Do you have DB connections available?

[mconfortino]

Yes. It is using 264 out of 500 avalaible. Also, storage is at 60%-80% on all six storage volumes. The system is responsive displaying montages, retrieving recordings, moving PTZ domes, etc. It has been running from almost ten months now, with a few reboots caused by power shortages. /var resides in its own volume, less than 40% full. I have scripts that delete events from the database and the filesystem when they are more than 45 days old, and another backups the database dayly and keeps the backups for two weeks. We also run a sanity check on the database every week. The only non-standard is that we disabled the posibility of deleting events for every user including admin. Even the trash can does not show. This brings some issues when there is a new version.

[dougmccrary]

Deja vu?

So a clone of a working camera didn't work even before you changed the ip for the moved one?

So a clone of a working camera didn't work even *before* you changed the ip for the moved one?

I know it seems unlikely, but this is sooo weird.

[mconfortino]

Yes it is! this is one of six city wide systems, ranging from 15 to 200 monitors. Cameras are Axis, Hikvision, Sony, Bosch, Pelco, Samsung, Vivotek, etc. Never has a problem like this.

[dougmccrary]

I think maybe it's time to start thinking hardware.
Maybe move some memory sticks around or replace to see if there's any change?

[iconnor]

Very strange. Without logs, cannot help.

A reminder that paid support is an option.

[iconnor]

Wait... over 100 cameras... so is this a multi-port install? Have you gone over the configured port range?