Bug 7361

Summary:	Fedora 20 in AB infrastructure: "Resource temporarily unavailable"
Product:	[Infrastructure] AutoBuilder	Reporter:	Beth Flanagan <elizabeth.flanagan>
Component:	autobuilder	Assignee:	Michael Halstead <mhalstead>
Status:	VERIFIED OBSOLETE	QA Contact:	Bogdan Alexandru Voiculescu <bogdanx.a.voiculescu>
Severity:	normal
Priority:	Medium	CC:	bogdanx.a.voiculescu, elizabeth.flanagan, infras.ab.watcher, Infras.watcher, meta.mr.watcher, meta.watcher, richard.purdie
Version:	unspecified
Target Milestone:	1.9
Hardware:	x86
OS:	arm
Whiteboard:
OS type for building Yocto:	---	Type of Regression:	---
Verified:		Documentation change:	Don't know

Description Beth Flanagan 2015-02-25 13:34:52 UTC

From: https://autobuilder.yoctoproject.org/main/builders/nightly-arm-lsb/builds/206/steps/BuildImages/logs/stdio

NOTE: recipe gcc-cross-arm-4.9.2-r0: task do_compile: Started
ERROR: Error executing a python function in /home/pokybuild/yocto-autobuilder/yocto-worker/nightly-arm-lsb/build/meta/recipes-core/glibc/glibc-locale_2.21.bb:

The stack trace of python calls that resulted in this exception/failure was:
File: 'split_and_strip_files', lineno: 216, function: <module>
     0212:    # End of strip
     0213:    #
     0214:
     0215:
 *** 0216:split_and_strip_files(d)
     0217:
File: 'split_and_strip_files', lineno: 209, function: split_and_strip_files
     0205:            sfiles.append((file, elf_file, strip))
     0206:        for f in kernmods:
     0207:            sfiles.append((f, 16, strip))
     0208:
 *** 0209:        oe.utils.multiprocess_exec(sfiles, oe.package.runstrip)
     0210:
     0211:    #
     0212:    # End of strip
     0213:    #
File: '/home/pokybuild/yocto-autobuilder/yocto-worker/nightly-arm-lsb/build/meta/lib/oe/utils.py', lineno: 170, function: multiprocess_exec
     0166:    pool = bb.utils.multiprocessingpool(nproc, init_worker)
     0167:    imap = pool.imap(function, commands)
     0168:
     0169:    try:
 *** 0170:        res = list(imap)
     0171:        pool.close()
     0172:        pool.join()
     0173:        results = []
     0174:        for result in res:
File: '/home/pokybuild/yocto-autobuilder/yocto-worker/nightly-arm-lsb/build/bitbake/lib/bb/utils.py', lineno: 916, function: wrap
     0912:    # Deal with a multiprocessing bug where signals to the processes would be delayed until the work
     0913:    # completes. Putting in a timeout means the signals (like SIGINT/SIGTERM) get processed.
     0914:    def wrapper(func):
     0915:        def wrap(self, timeout=None):
 *** 0916:            return func(self, timeout=timeout if timeout is not None else 1e100)
     0917:        return wrap
     0918:    multiprocessing.pool.IMapIterator.next = wrapper(multiprocessing.pool.IMapIterator.next)
     0919:
     0920:    return multiprocessing.Pool(*args, **kwargs)
File: '/usr/lib64/python2.7/multiprocessing/pool.py', lineno: 655, function: next
     0651:
     0652:        success, value = item
     0653:        if success:
     0654:            return value
 *** 0655:        raise value
     0656:
     0657:    __next__ = next                    # XXX
     0658:
     0659:    def _set(self, i, obj):
Exception: OSError: [Errno 11] Resource temporarily unavailable

ERROR: Function failed: split_and_strip_files
ERROR: Logfile of failure stored in: /home/pokybuild/yocto-autobuilder/yocto-worker/nightly-arm-lsb/build/build/tmp/work/armv5e-poky-linux-gnueabi/glibc-locale/2.21-r0/temp/log.do_package.10029
NOTE: recipe glibc-locale-2.21-r0: task do_package: Failed
ERROR: Task 453 (/home/pokybuild/yocto-autobuilder/yocto-worker/nightly-arm-lsb/build/meta/recipes-core/glibc/glibc-locale_2.21.bb, do_package) failed with exit code '1'

Comment 1 Beth Flanagan 2015-03-05 15:44:50 UTC

I'm pretty sure at this point that this is an issue with the Fedora 20 machine's hardware.

1. We removed f20 from the pool and have not seen this since.
2. I'm running f20 and have never seen these errors.

I'm going to kick this over to halstead and see if he has some thoughts.

Comment 2 Richard Purdie 2015-03-12 12:51:45 UTC

Michael: Any update on whether this was some kind of hardware issue?

Comment 3 Michael Halstead 2015-03-18 10:44:02 UTC

All hardware tests I ran passed without any trouble and there is nothing in the logs. If you ssh to fedora20.yp you can see a troublesome make (pid 12464) still running from the last failure. Perhaps it can offer clues?

Comment 4 Michael Halstead 2015-05-18 15:45:29 UTC

At this point I propose we reload the machine with Fedora22 which is due out next week. It doesn't seem this will be looked into more deeply and this machine is currently on standby.

Any objections?

Comment 5 Beth Flanagan 2015-06-04 17:01:40 UTC

That sounds fine. Let me know when it's done.

Comment 6 Stephen K Jolley 2015-06-08 19:53:28 UTC

It appears the question was answered.

Comment 7 Michael Halstead 2015-06-08 20:43:07 UTC

Wiped and starting over with Fedora 22.

Comment 8 Bogdan Alexandru Voiculescu 2015-06-17 08:30:05 UTC

Nothing to be verified.