We upgraded to 7.0.0.50 yesterday and it seems i can’t connect to the Slave Logs anymore from the Monitor (both Slave and Task panels affected). Opening the Log on the Slave itself (via RDP connection) is no problem.
Will any log help? If so, which one?
Cheers,
Holger
Hi Holger,
I can’t seem to reproduce this problem here. Are you getting an error in the slave log window that pops up?
Cheers,
Ryan
Actually,… that’s the thing. The Slave Log window doesn’t even open. I tried this on different machines: nada.
Only the machine that is running Pulse it works. That machine is actually a Windows XP 64bit system (i know, basically forbidden 
) in a VM. But that shouldn’t really make any difference, should it?
Cheers,
Holger
Hmm, interesting. Can you try running the client installer manually on a machine where this isn’t working to see if that helps? Maybe something went wrong during the auto upgrade…
No auto-upgrade in place here at all. For now, i’m running silent installers on all machines.
Ah, cool.
Can you open the Task Manager on a machine that is having this problem and see if there are already any deadlinecommand processes running? If there are, can you try killing them and then try connecting again? Maybe you could try restarting the machine too?
The only ‘deadline…’ processes that are running are deadlinemonitor.exe and deadlinelauncher.exe. Also, between 1 to 3 reboots have already happened since i first saw this issue last night. Basically, right after the upgrade.
Thanks!
If you open the Console panel in the Monitor, are you seeing any error messages when you try to connect? If so, can you post them?
If not, let’s try manually connecting from the command line. Open a command prompt on the Monitor’s machine, and run this command:
"C:\Program Files\Thinkbox\Deadline7\bin\deadlinecommandbg.exe" -InternalConnectToSlaveLog "SLAVENAME" -1 "REPOSITORYPATH"If your slave name is “NODE05” and your repository path is “\server\DeadlineRepository7”, the command would look like this:
"C:\Program Files\Thinkbox\Deadline7\bin\deadlinecommandbg.exe" -InternalConnectToSlaveLog "NODE05" -1 "\\server\DeadlineRepository7"I’m expecting the UI to not show up, and hopefully for an error message to get printed to the command prompt window. If this happens, please post the error message.
Thanks!
Ryan
Sorry to fall short of your expectations here. The command line method actually works.
When i tried with the Monitor all i got in the Console panel was this:
2014-12-04 17:24:17: Connecting to slave log: CELL-RS-10
and no window opened up.
Hmm… well that’s unexpected.
One last thing to test - can you launch a slave on the Monitor machine, and then see if:
Thanks!
Ryan
Well, I’m quite stumped at this point, especially since we didn’t change anything with respect to the slave logs in any of the release candidates.
One last thing that’s probably worth trying is to uninstall the Deadline client and then reinstall it again.
Well… guess i’m gonna try that then
Will test this on one machine first and then see if it helps.
I’ll let you know.
Ok. I just uninstalled and then re-installed. No luck, unfortunately.
Darn…
We’ve tried to reproduce this on a hand full of machines here (with various operating systems), and everything is working just fine. Even if a connection can’t be established with the remote slave, the window still pops up.
So is every machine except your pulse machine affected by this? Can the slave machines connect to the logs of other slaves if you launch the monitor on them? I really can’t think of a reason why this would happen, especially if the same command from a cmd prompt works just fine…
On all the machines i’ve checked sof far this happen. except for the Pulse machine. To be able to give you a more precise picture i guess i have to check quite a few more. Will try to do so asap and give you an update then.
Not sure if this’ll help in any way but i just had Process Monitor from the Sysinternals Suite running while trying to open a Slave log. I’m by far no expert with this but there seem to be some error messages in there which are related to the issue. I saved the log in all the formats it exports (csv, native and xml) and attached it.
E.g. at 22:08:28,1372280 it says “Buffer Overflow” for the command
"C:\Program Files\Thinkbox\Deadline7\bin\deadlinecommandbg.exe" -InternalConnectToSlaveLog "CELL-RS-12" -1 "\\cell-dead-01\deadlinerepo\"Maybe this gives you any pointers? FYI, i had the filter set to only log and display events that contain “deadline” in their process name.
It’s probably also worth noting that i didn’t do any Windows updates right before or since the Deadline update.
procmon_logs_2014-12-04.zip (538 KB)
I’ll take a look through the logs.
So if you run this exact command from a command prompt on the same machine, it works right?
"C:\Program Files\Thinkbox\Deadline7\bin\deadlinecommandbg.exe" -InternalConnectToSlaveLog "CELL-RS-12" -1 "\\cell-dead-01\deadlinerepo\"I googled buffer overflow messages in procmon, and found this:
superuser.com/questions/491597/p … r-overflow
So those messages might just be completely normal…
This is weird. I now tried this around 5 to 6 times in a DOS box with different machine names and it doesn’t work at all now. It definitely worked when you first asked me to try this. I also don’t get any error message on the command line nor any output, just an empty line and no Slave log window.