I would like to stress again, that it would be very important to evaluate performance of the monitor with 15k-20k+ jobs. We constantly hover over 10.000, but average around 13-14k. The monitor becomes so sluggish, its impractical for everyday use. The only remedy we have is to only keep jobs around for a couple of days, which then hurts us on the production side, as we frequently need older jobs.
It becomes a management nightmare for both production, 200+ artists, pipeline etc., simply due to the monitor gui being slow with the job count we have.
We are yet to see how deadline7 performs with such a job count, but i’m hoping you guys are doing some profiling on your end with such job counts. We are currently at 17000+ and we only expect that to grow.
We have put significant effort into improving some aspects of the Monitor’s performance in Deadline 7, like how the filtering system works, and we’ll continue to try and improve performance whenever possible. We’ll definitely run some more tests with 20,000 jobs to see how things perform.
Note that slow down can still occur though if there are lots of panels open, so we recommend to store different panel layouts as pinned layouts, and switch between them as necessary to keep the panel count lower.
Cheers,
Ryan
Currently the deadline7 monitor seems slower than the deadline6 one to be honest. If i leave it open for a couple of days, it very near becomes unusable. We have 20 jobs in there, with 200 slaves. Not sure whats going on… qt5 issue?
Literally takes 10 seconds to select a job, moving dividers is extremely sluggish…
What’s the memory and cpu usage look like?
Also, could you try disabling the Grup Jobs By Batch Name option in the Monitor Options to see if that helps? Note that you will have to restart the Monitor afterwards.
Regardless, we’ll see if we can reproduce this slowdown here.
Cheers,
Ryan
Memory usage is very low, around 57megs. I dont see any cpu usage unless im doing any activity, in which case it spykes to 100% on a core on operations like dragging dividers.
Related note: as i was clicking around on random jobs, ram usage went up to 149 megs and stayed there. I can keep increasing ram usage by stepping between jobs. Depending on the number of tasks in the jobs, it goes up by 100k-1meg every time i step on a job.
The setting you mentioned requires a monitor restart, so its not going to be representative… After a restart the monitor is always happy.
Btw, i turned off that option, and have not restarted the monitor. Some of the graphs went nuts, showing no users/groups:
Restarting the Monitor will fix that. 
