Smedge and Muster

[Dman3d]

anyone use smedge or muster? any reason to consider over backburner (other than backburner being a general POS)

[Adam H. Stewart]

Didn't know Smedge did Max

Are you looking at Smedge for Maya ? I didn't know it did Max.
We use Deadline and have no complaints. Once you have it set up correctly it is really nice - you can also run a huge number of programs like, Combustion, AE, Lightwave - although we don't have it set up for all that many just Max and Maya - it is easy to use and very reliable.

[vance3d]

we also use Deadline and I can highly recommend it. we use it for max and after effects.

V Miller

[studioDIM]

I had smedge doing all sorts of things with maya and Mr, a while back.
Was glad i dropped it, but wouldn't know if it got any better or was indeed a configuration issue...

Lele

[Morbid Angel]

dont know about smedge but we use muster, and dont! it doesnt support many things which are basic and required.

[Dman3d]

Thanks, I will look up deadline

[Morbid Angel]

downside about deadline is that it costs 2g / piece

[Dman3d]

The question then is is that is the manager how many slaves does that support?

[Morbid Angel]

no idea... why backburner doesnt do it for you?
You see, programs like muster as an example, they do not support direct submission from max/maya to it. Therefore you cant specify time range and output and click network render as if you would with backburner. You have to manually go into muster, navigate to your scene, then navigate to your output path, specify frame range etc...its all very confusing and calls for big errors from the user. And if you accidently click on another job your settings will dissapear and you have to start all over.
Not to mention that muster doesnt support frame blocks. For example if you wanted to submit frames 13-25,50-123, 220,230,235
muster does not have an option for that. So the only thing you can do is to manually load each of the blocks.
That sux big time.

[Dman3d]

Well good info. I have no intention of migrating on the fly merely looking
around.

As for backburner we call it assburner and call our render managers
render manglers if that gives an indication.

The single biggest problem is the manager crashing.
We run two managers they are split up for group usage and for the
idea that perhaps assburner, sorry, habit, backburner will run smoother
and hopefully not crash.

We run between 30 to 40 slaves full time, lol depending on which ones
are up. There seems to be a good deal of crashing when we get
over 30 slaves on one manager and/or there are alot of people using
the monitor.

We have recenlty put a policy in place to use a couple walk up machines
for the monitor and no one is to run it local, at least until we can figure
out why it is so unstable. We have been looking for answers forever
and had no solutions.

I have been asking everyone to use notifications for what seems like forever
to prevent too many people connecting.

The machines we run the manglers on are usually last years work stations
1CPU and 1GB RAM with 10/100 NIC. I say usally as one of them just
ate crap friday and had to be moved to another box, a hardware issue.

Now the other problem is people blame backburner for stuff that has nothing
to do with backburner so it has a bad rep around our shop.
To the point of "Had a flat tire this morning, must have ran over backburner."

Anyone have advice on Backburner or more logical ways of trouble shooting manager crashes and the like?

Oh there is one thing we have learned. If backburner crashes you will
find an extra job or more in the job folder that does not appear in the
manager itself. You have to delete these phantom/rogue jobs or it will
crash again shortly after restarting it.

Also to back up the idea of less people using the monitor is better the
second manager or better to say which ever manager is not using the most
servers and thus has less if not no people connecting to it never crashes
(of course that is somewhat blanket but I honestly cannot recall a crash
on mangler2)

Sorry cant keep my pinky from the hard return

What about networking, does anyone do anything special in terms of slaves
like keeping them on a separate hub or branch? I think our IT set it up
like that but I cant remember.

[Morbid Angel]

sounds pretty serious. For us, backburner never crashes. Nor does the monitor. Its odd really. Perhaps you should contact autodesk support for this. Since this is not normal.

[Dman3d]

we will try again I suppose, we get this non-stop, well had it non-stop
for years. Any time we try to get help it deadends, almost like
backburner was written by no one and it does not exist.

But fair point aye, its time for another go at backburner. Our current
IT guys is pretty sharp and cooperative so I will start the engines
up again.

Anyone else out ther run backburner in a similar way or environment?

quote the cannibal corpse "...make them suffer...."

[percydaman]

ya our backburner gives problems on a semi regular basis...

[Dman3d]

you know I had forgotten about this. I will check tomorrw when I am sure
I will have to go in due to a drop crapping out.

http://www.chaosgroup.com/forum/phpB...ht=connections

Not sure what OS the manager is on at this point

I am not clear on how many connections Server 2003 allows after reading
that and our old IT guy never new. I will check with out new IT guy.
I know he dislikes linux, not sure why, but I remember reading this
before and recall the samba combo seemed like a good idea

Anyone have deeper info or links about setting up a manager on Linux?

Im sure we can sort it but I really need and want to resolve our unstable
manager issue asap, had enough of it.

We have backburner set up to view over the web which is handy from
home, at least I know when a job goes south.

[Dynedain]

Make sure you actually have max installed on the backburner manager... stupid, I know, but it solves a lot of issues.

It sounds like running two managers probably causes conflicts as well.

If you want to make sure people aren't using the monitor locally, just remove it from their machines, and leave it only on the 2-3 machines you want people using it on.

We have 14 nodes and 2 workstations using backburner (manager is on a Win2k Server box) and don't have any crashing problems related to backburner. The manager is even an old P3 and handles a bunch of other server tasks (like nightly backups, the domain controller, and keysever for a couple of apps).