Wolrajh

Wolrajh

About Me

Expertise
Not Specified
Location
Not Specified
Website

Connect

Recent Forum Posts

deadline "socket.gaierror: [Errno 11001] getaddrinfo failed" April 19, 2019, 9:01 a.m.

Ah yes, I see. I can't go static IP, especially these days with a massive revamp of our network, but modifying the ‘s’ argument directly in the CommandLine Settings in the Deadline job made it work. Just replaced my hostname with my actual IP on the network ( kept the port ) and it went smoothly.

Obviously it can't be final, but at least it worked !

deadline "socket.gaierror: [Errno 11001] getaddrinfo failed" April 18, 2019, 11:05 a.m.

Either Errno 11001 or Errno 11004, actually.

Most of the latest additions and corrections to PDG seem to have taken care of the submission errors we were getting, as well as simplify the whole process. No need to point to the repository, HFS is correctly exploited directly from within Houdini, it's waaay better.

However, I'm now getting a weird error : upon submitting to deadline via PDG, the job somehow fails to, well, tell it's actually finished.
The result of a simple ROP correctly writes on the disk, but the job starts looping with an error, and Houdini in turn can't mark the tasks as complete, as it is still waiting on the deadline status report.

From what I understand from that error, Deadline can't resolve a server address, but, as a lowly noobie, I'm not sure what server it actually tries to access here. I'm pretty sure this is an issue with our network, but understanding the error could help us find a resolution.

Maybe it's just another variable to pass through from Houdini ?

Extract from the Deadline log :

2019-04-18 16:46:29:  0: STDOUT: Loading .hip file VPATH/PDG_deadline_test/PDG_deadline_test.hip.
2019-04-18 16:46:31:  0: STDOUT: Traceback (most recent call last):
2019-04-18 16:46:31:  0: STDOUT: PDG_RESULT: ropgeometry1_ropfetch1_3;-1;'__PDG_DIR__/geo/PDG_deadline_test.ropgeometry1.3.bgeo.sc';;0
2019-04-18 16:46:31:  0: STDOUT:   File "V:/PATH/PDG_deadline_test/pdgtemp/40756/scripts/rop.py", line 502, in <module>
2019-04-18 16:46:31:  0: STDOUT:     cooker.cookSingleFrame(args)
2019-04-18 16:46:31:  0: STDOUT:   File "V:/PATH/PDG_deadline_test/pdgtemp/40756/scripts/rop.py", line 203, in cookSingleFrame
2019-04-18 16:46:31:  0: STDOUT:     reportResultData(parm.evalAtFrame(args.start), server_addr=args.server)
2019-04-18 16:46:31:  0: STDOUT:   File "V:\PATH\PDG_deadline_test\pdgtemp\40756\scripts\pdgcmd.py", line 222, in reportResultData
2019-04-18 16:46:31:  0: STDOUT:     result_data_tag, hash_code, jobid)
2019-04-18 16:46:31:  0: STDOUT:   File "\\config.int.mtc.fr\config\_MTC_C~1\_CONFI~1\HOUDIN~1.229\python27\lib\xmlrpclib.py", line 1243, in __call__
2019-04-18 16:46:31:  0: STDOUT:     return self.__send(self.__name, args)
2019-04-18 16:46:31:  0: STDOUT:   File "\\config.int.mtc.fr\config\_MTC_C~1\_CONFI~1\HOUDIN~1.229\python27\lib\xmlrpclib.py", line 1602, in __request
2019-04-18 16:46:31:  0: STDOUT:     verbose=self.__verbose
2019-04-18 16:46:31:  0: STDOUT:   File "\\config.int.mtc.fr\config\_MTC_C~1\_CONFI~1\HOUDIN~1.229\python27\lib\xmlrpclib.py", line 1283, in request
2019-04-18 16:46:31:  0: STDOUT:     return self.single_request(host, handler, request_body, verbose)
2019-04-18 16:46:31:  0: STDOUT:   File "\\config.int.mtc.fr\config\_MTC_C~1\_CONFI~1\HOUDIN~1.229\python27\lib\xmlrpclib.py", line 1311, in single_request
2019-04-18 16:46:31:  0: STDOUT:     self.send_content(h, request_body)
2019-04-18 16:46:31:  0: STDOUT:   File "\\config.int.mtc.fr\config\_MTC_C~1\_CONFI~1\HOUDIN~1.229\python27\lib\xmlrpclib.py", line 1459, in send_content
2019-04-18 16:46:31:  0: STDOUT:     connection.endheaders(request_body)
2019-04-18 16:46:31:  0: STDOUT:   File "\\config.int.mtc.fr\config\_MTC_C~1\_CONFI~1\HOUDIN~1.229\python27\lib\httplib.py", line 1038, in endheaders
2019-04-18 16:46:31:  0: STDOUT:     self._send_output(message_body)
2019-04-18 16:46:31:  0: STDOUT:   File "\\config.int.mtc.fr\config\_MTC_C~1\_CONFI~1\HOUDIN~1.229\python27\lib\httplib.py", line 882, in _send_output
2019-04-18 16:46:31:  0: STDOUT:     self.send(msg)
2019-04-18 16:46:31:  0: STDOUT:   File "\\config.int.mtc.fr\config\_MTC_C~1\_CONFI~1\HOUDIN~1.229\python27\lib\httplib.py", line 844, in send
2019-04-18 16:46:31:  0: STDOUT:     self.connect()
2019-04-18 16:46:31:  0: STDOUT:   File "\\config.int.mtc.fr\config\_MTC_C~1\_CONFI~1\HOUDIN~1.229\python27\lib\httplib.py", line 821, in connect
2019-04-18 16:46:31:  0: STDOUT:     self.timeout, self.source_address)
2019-04-18 16:46:31:  0: STDOUT:   File "\\config.int.mtc.fr\config\_MTC_C~1\_CONFI~1\HOUDIN~1.229\python27\lib\socket.py", line 557, in create_connection
2019-04-18 16:46:31:  0: STDOUT:     for res in getaddrinfo(host, port, 0, SOCK_STREAM):
2019-04-18 16:46:31:  0: STDOUT: socket.gaierror: [Errno 11001] getaddrinfo failed
2019-04-18 16:46:31:  0: INFO: Process exit code: 1

FfmpegEncodeVideo bug ? April 10, 2019, 7:40 a.m.

Kind of bumping myself here, but I'm curious to know if this is still in the plans ; right now the best solution we found was to change the working directory on the localsheduler to where the images are stored ( eg Wimages ), duplicate that localscheduler, and tell the topnet to use it as the default scheduler.

Just changing the path on the initial localscheduler does not seem to work as it will still believes the pdg dir is on our drive with the scene ( eg Vscenes ) and try to use the filenamelist from V, until we restart Houdini. Duplicating and switching localschedulers out on the other will force the topnet to evaluate from W without having to reset the Houdini session.

Or we could edit the python and add some variables to make it save the filelist on whatever drive we need, but that's a little contrived

So obviously it's still very clunky, hence my little bump on that particular topic !