We had a couple of artists report that their jobs’ statuses show out of date states for extended periods of time, in the range 10-15+ mins before updating to the right status. As you can tell, we have some people glued to the monitors who like to watch paint dry 
The statuses do eventually show up as correct. We had the job status updates set to 30seconds, i 've reduced it to 10 now, but im not sure that would help in this case. Afaik the artist, and the slaves rendering his job were all of the latest build.
Is it possible to email us directly the next time this happens, and pass along the job ID? We can then take a look at your db and see what the timestamp looks like for the job. That will help us determine if the issue is with the timestamp in the db or with the way the monitor is querying it.
Thanks!
Sure, you could log in now actually, there are a couple like that.
Look for only queued&rendering, and only the nuke group. Any job older than an hour there still in rendering/queued state is like that.
Although if you start a fresh monitor, it might be ‘ok’…
Yup, found a few, and their timestamps were fine, which means the Monitor missed them. Restarting the monitor of course fixed it. We need to either figure out a way to prevent the monitor from missing these updates, or have a way for the Monitor to “reset” it’s internal job timestamp so that it can reload things it may have missed.
Thanks!
Ryan
I would vote for the refresh button to do that, and maybe have it reset periodically too?