test_site = TestSite(self,site_spec)
test_slice=TestSlice(self,test_site,slice_spec)
tasks += slice_method (test_slice, self.options)
- return Completer (tasks).run (decorator_self.timeout, decorator_self.silent, decorator_self.period)
+ return Completer (tasks, message=method.__name__).run (decorator_self.timeout, decorator_self.silent, decorator_self.period)
# restore the doc text from the TestSlice method even if a bit odd
wrappee.__name__ = method.__name__
wrappee.__doc__ = slice_method.__doc__
'sfa_rspec_empty@1', 'sfa_allocate_empty@1', 'sfa_provision_empty@1','sfa_check_slice_plc_empty@1', SEPSFA,
'sfa_delete_slice@1', 'sfa_delete_user@1', SEPSFA,
'cross_check_tcp@1', 'check_system_slice', SEP,
+ # for inspecting the slice while it runs the first time
+ #'fail',
# check slices are turned off properly
'empty_slices', 'ssh_slice_off', 'slice_fs_deleted_ignore', SEP,
- # xxx for Thomas
- 'fail',
# check they are properly re-created with the same name
'fill_slices', 'ssh_slice_again', SEP,
'gather_logs_force', SEP,
utils.header("checking nodes boot state (expected %s)"%target_boot_state)
tasks = [ CompleterTaskBootState (self,hostname) \
for (hostname,_) in self.all_node_infos() ]
- return Completer (tasks).run (timeout, graceout, period)
+ message = 'check_boot_state={}'.format(target_boot_state)
+ return Completer (tasks, message=message).run (timeout, graceout, period)
def nodes_booted(self):
return self.nodes_check_boot_state('boot',timeout_minutes=30,silent_minutes=28)
return True
# probing nodes
- def check_nodes_ping(self,timeout_seconds=120,period_seconds=10):
- class CompleterTaskPingNode (CompleterTask):
+ def check_nodes_ping(self, timeout_seconds=30, period_seconds=10):
+ class CompleterTaskPingNode(CompleterTask):
def __init__ (self, hostname):
self.hostname=hostname
- def run(self,silent):
+ def run(self, silent):
command="ping -c 1 -w 1 %s >& /dev/null"%self.hostname
return utils.system (command, silent=silent)==0
def failure_epilogue (self):
period=timedelta (seconds=period_seconds)
node_infos = self.all_node_infos()
tasks = [ CompleterTaskPingNode (h) for (h,_) in node_infos ]
- return Completer (tasks).run (timeout, graceout, period)
+ return Completer (tasks, message='ping_node').run (timeout, graceout, period)
# ping node before we try to reach ssh, helpful for troubleshooting failing bootCDs
def ping_node (self):
"Ping nodes"
return self.check_nodes_ping ()
- def check_nodes_ssh(self,debug,timeout_minutes,silent_minutes,period_seconds=15):
+ def check_nodes_ssh(self, debug, timeout_minutes, silent_minutes, period_seconds=15):
# various delays
timeout = timedelta(minutes=timeout_minutes)
graceout = timedelta(minutes=silent_minutes)
vservername=self.vservername
if debug:
message="debug"
+ completer_message = 'ssh_node_debug'
local_key = "keys/%(vservername)s-debug.rsa"%locals()
else:
message="boot"
+ completer_message = 'ssh_node_boot'
local_key = "keys/key_admin.rsa"
utils.header("checking ssh access to nodes (expected in %s mode)"%message)
node_infos = self.all_node_infos()
- tasks = [ CompleterTaskNodeSsh (nodename, qemuname, local_key, boot_state=message) \
+ tasks = [ CompleterTaskNodeSsh (nodename, qemuname, local_key,
+ boot_state=message, dry_run=self.options.dry_run) \
for (nodename,qemuname) in node_infos ]
- return Completer (tasks).run (timeout, graceout, period)
+ return Completer (tasks, message=completer_message).run (timeout, graceout, period)
def ssh_node_debug(self):
"Tries to ssh into nodes in debug mode with the debug ssh key"
test_node = TestNode (self,test_site,node)
test_sliver = TestSliver (self, test_node, test_slice)
tasks.append ( CompleterTaskInitscript (test_sliver, stamp))
- return Completer (tasks).run (timedelta(minutes=5), timedelta(minutes=4), timedelta(seconds=10))
+ return Completer (tasks, message='check_initscripts').run (timedelta(minutes=5), timedelta(minutes=4), timedelta(seconds=10))
def check_initscripts(self):
"check that the initscripts have triggered"
for spec in specs:
port = spec['port']
# server side
- s_test_sliver = self.locate_sliver_obj_cross (spec['server_node'],spec['server_slice'],other_plcs)
- if not s_test_sliver.run_tcp_server(port,timeout=20):
- overall=False
+ # the issue here is that we have the server run in background
+ # and so we have no clue if it took off properly or not
+ # looks like in some cases it does not
+ s_test_sliver = self.locate_sliver_obj_cross (spec['server_node'], spec['server_slice'], other_plcs)
+ if not s_test_sliver.run_tcp_server(port, timeout=20):
+ overall = False
break
# idem for the client side
- c_test_sliver = self.locate_sliver_obj_cross (spec['client_node'],spec['client_slice'],other_plcs)
+ c_test_sliver = self.locate_sliver_obj_cross (spec['client_node'], spec['client_slice'], other_plcs)
# use nodename from locatesd sliver, unless 'client_connect' is set
if 'client_connect' in spec:
destination = spec['client_connect']
else:
- destination=s_test_sliver.test_node.name()
- if not c_test_sliver.run_tcp_client(destination,port):
- overall=False
+ destination = s_test_sliver.test_node.name()
+ if not c_test_sliver.run_tcp_client(destination, port):
+ overall = False
return overall
# painfully enough, we need to allow for some time as netflow might show up last
period = timedelta (seconds=period_seconds)
tasks = [ CompleterTaskSystemSlice (test_node, self.options.dry_run) \
for test_node in self.all_nodes() ]
- return Completer (tasks) . run (timeout, silent, period)
+ return Completer (tasks, message='_check_system_slice') . run (timeout, silent, period)
def plcsh_stress_test (self):
"runs PLCAPI stress test, that checks Add/Update/Delete on all types - preserves contents"