diff options
Diffstat (limited to 'src/ceph/qa/tasks/die_on_err.py')
-rw-r--r-- | src/ceph/qa/tasks/die_on_err.py | 70 |
1 files changed, 0 insertions, 70 deletions
diff --git a/src/ceph/qa/tasks/die_on_err.py b/src/ceph/qa/tasks/die_on_err.py deleted file mode 100644 index bf422ae..0000000 --- a/src/ceph/qa/tasks/die_on_err.py +++ /dev/null @@ -1,70 +0,0 @@ -""" -Raise exceptions on osd coredumps or test err directories -""" -import contextlib -import logging -import time -from teuthology.orchestra import run - -import ceph_manager -from teuthology import misc as teuthology - -log = logging.getLogger(__name__) - -@contextlib.contextmanager -def task(ctx, config): - """ - Die if {testdir}/err exists or if an OSD dumps core - """ - if config is None: - config = {} - - first_mon = teuthology.get_first_mon(ctx, config) - (mon,) = ctx.cluster.only(first_mon).remotes.iterkeys() - - num_osds = teuthology.num_instances_of_type(ctx.cluster, 'osd') - log.info('num_osds is %s' % num_osds) - - manager = ceph_manager.CephManager( - mon, - ctx=ctx, - logger=log.getChild('ceph_manager'), - ) - - while len(manager.get_osd_status()['up']) < num_osds: - time.sleep(10) - - testdir = teuthology.get_testdir(ctx) - - while True: - for i in range(num_osds): - (osd_remote,) = ctx.cluster.only('osd.%d' % i).remotes.iterkeys() - p = osd_remote.run( - args = [ 'test', '-e', '{tdir}/err'.format(tdir=testdir) ], - wait=True, - check_status=False, - ) - exit_status = p.exitstatus - - if exit_status == 0: - log.info("osd %d has an error" % i) - raise Exception("osd %d error" % i) - - log_path = '/var/log/ceph/osd.%d.log' % (i) - - p = osd_remote.run( - args = [ - 'tail', '-1', log_path, - run.Raw('|'), - 'grep', '-q', 'end dump' - ], - wait=True, - check_status=False, - ) - exit_status = p.exitstatus - - if exit_status == 0: - log.info("osd %d dumped core" % i) - raise Exception("osd %d dumped core" % i) - - time.sleep(5) |