Logon failure: unknown user name or bad password.

[connnor.v]

I was running through some jobs on our farm last night when I started seeing this error:

requesting work for: 59087.2
got work: 59087:121-133 - running
Logon failure: unknown user name or bad password.
ERROR: exit status: 1

Looks like it was isolated to two machines that were running on the farm, although nothing appeared to be wrong with them at all.

Can you please explain what this error means, and where it arises?

Additionally, earlier in the night last night I spotted a "badlogin" state on a different machine. Did that have anything to do with this error? Why do "badlogins" happen?

[jburk]

These are 2 different problems.

badlogin

The badlogin job state occurs when the worker is running in user mode instead of proxy mode, and the worker is unable to authenticate as the job's user.  That user will need to inform Qube of their Windows password. 

See http://docs.pipelinefx.com/pages/viewpage.action?pageId=1867833

for a detailed description.

But I will mention that running a Window worker in user mode rather than proxy mode is very rare, and is usually the result of a misconfiguration of the worker

http://docs.pipelinefx.com/display/QUB064/Authentication+Modes+when+in+Service+Mode%3A+user+and+proxy


Logon failure: unknown user name or bad password.

This is when the worker tries to access a share on a file server that challenges for a username and password when accessed; it's not sufficient to be logged into a host in a Windows domain.  You've probably seen this when logging into a machine for the first time and trying to access a share on a non-Windows file server, or a server in a Windows Workgroup.  And you've probably checked the "cache my credentials" checkbox, so you only have to enter this info once.

Qube workers can do the same thing with the worker_drive_map parameter. 

http://docs.pipelinefx.com/display/QUB064/worker_drive_map