When I choose to requeue a Task I love that Deadline asks me if I want to Blacklist the slave from the job, I’d like to be able to say Blacklist from all my jobs as well…
Hello Dave,
While I think I understand what you are asking, I think the disable slave option would be what we would suggest using here. I can suggest adding a ‘disable slave’ option there if you’d like so that we can put it on our wishlist, allowing you to disable the slave. Just to help you understand my thinking on directing the idea this way is I am not sure if ‘blacklist for all jobs’ would be clear for some. Blacklist for only current jobs? If you’re blacklisting for all jobs, ever, then it’s basically disabling the slave. Does that make sense?
Yeah actually that does make more sense… we get blades that occasionally mess up, fill up their HDs or just need a rebuild.
Also to have a ‘Restart offending machines’ tickbox would be nice, as that often fixes whatever their issue is.
Perhaps it would be better to think about ways to detect and/or resolve the root cause of the slave instability? Better HD filling up warnings in monitor? A list of known scenarios would help to focus our thoughts here. Just throwing it out there 
Yeah definitely a warning for when the HDs fill up / option to clean up logical places ‘Bitmap Pager Cache’ Temporary folders…
We seem to be having a lot of blades getting stuck on the ‘starting-up’ stage of rendering at the moment, happens to lots of different machines and not every time. I’m wondering if it’s because we have a common folder for our plugins on the network. It seems like they get stuck loading plugins and seems again (not scientifically yet) to be on the smaller quicker to load jobs… Need to pin it down as it’s quite annoying!
Hmmm…I’ll have a think about this a little more…got a couple of ideas here…
You could write a pre job script which does some cleanup for you or even use one of our job or slave “event” callbacks to get the slave or slave de-queuing a job to search various locations and do some house-keeping prior to starting a job or perhaps, whenever a slave first boots up. Various options here. 
Yeah, there is a number of ‘things’ this could be and I’ve been on a bit of mission late last year and earlier this year to finally nail it. Off the top of my head, depending on 3dsMax version/ext/sp, it could be Comms Center, Scene States, Scene Explorer, Sequencer, running as a service OR any combo of each of those listed. I’m pretty sure, it’s not your plugins being located centrally that is causing this issue. I would need to see quite a few (50+) of your job logs reports to be able to narrow it further down than this, as a spread across multiple jobs/slaves all displaying any error, so that I could try and visualise any pattern. If it helps at all, I’m pretty sure I have everything nailed in v7.1 beta 1 (which is now up to beta 4). Various fixes also appeared in earlier versions - v6.2.1.50, v7.0.0.54, v7.0.2.3 and finally, the big daddy, v7.1. All the fixes introduced in each version were worthy additions, but it’s not until the final v7.1, that I’m 99.999% sure I have every base covered. Then again, I have been wrong before… 
Yeah I think we’ll upgrade to Deadline 7.1 and hope it fixes it! Just got to magic the time from somewhere!