ppigazzini

The first issue could be due to my bug slip, with all fleets configured to restart if the worker is stopping for a fatal exception. The workers stopped being rejected...

``` sudo journalctl -u fishtest@6543 --since "30 days ago" | grep "is already a worker" | wc -l 652 ``` Mostly "ChessDBCN" and "technologov", but there are at least a...

> @ppigazzini Do you see other duplication warnings for this worker "ChessDBCN-11cores-ccd4f156" outside server restarts? ```yaml sudo journalctl -u fishtest@6543 --since "60 days ago" | grep "is already a worker"...

IMO the issue is triggered in some way by a lan connection loss/retard, workers side (for "ChessDBCN" and "technologov") or VPS side (for other users). The issue of the 24...

The server temporary rejects 23-26 duplicated workers/day, perhaps on Discord people is overreacting a bit. ``` sudo journalctl -u fishtest@6543 --since "180 days ago" | grep "is already a worker"...

The bug killed the worker when requesting a new task, some code auto restarted the worker. The duplicated worker issue of the 24 April disappeared after merging the bugfix.

Unfortunately, this should explain the low core count... 

DEV updated.

```yaml Jul 04 23:30:22 dfts-0 pserve[102834]: Traceback (most recent call last): Jul 04 23:30:22 dfts-0 pserve[102834]: File "/home/usr00/fishtest/server/env/lib/python3.12/site-packages/pyramid_mako/__init__.py", line 148, in __call__ Jul 04 23:30:22 dfts-0 pserve[102834]: result = template.render_unicode(**system)...

The worker (python 3.6.x) stops, wrong url, missing the repo info. (all urls before the wrong one are fine). - PR, wrong: ```bash Running master vs master Variable task sizes...