ansible/community.general
1
0
Fork 0
mirror of https://github.com/ansible-collections/community.general.git synced 2024-09-14 20:13:21 +02:00
Code Releases Activity

Check status of finished spot instance requests (#4990)

Browse source 
Per #3877, the code to wait for spot instance requests to finish would
hang for the full wait time if any spot request failed for any reason.
This commit introduces status checks for spot requests, so if the
request fails, finishes, or is cancelled the task will fail/succeed
accordingly.

One edge case introduced here is tha if a user terminates the instance
associated with the request manually it won't fail the play, under the
presumption that the user *wants* the instance terminated.
This commit is contained in:
Ryan Brown 2016-09-30 11:26:11 -04:00 committed by Matt Clay
parent 22d29be478
commit 135b92bf4e
1 changed files with 58 additions and 17 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

75
lib/ansible/modules/cloud/amazon/ec2.py
View file

@ -793,6 +793,63 @@ def boto_supports_param_in_spot_request(ec2, param):
method = getattr(ec2, 'request_spot_instances')
return param in method.func_code.co_varnames
def await_spot_requests(module, ec2, spot_requests, count):
"""
Wait for a group of spot requests to be fulfilled, or fail.
module: Ansible module object
ec2: authenticated ec2 connection object
spot_requests: boto.ec2.spotinstancerequest.SpotInstanceRequest object returned by ec2.request_spot_instances
count: Total number of instances to be created by the spot requests
Returns:
list of instance ID's created by the spot request(s)
"""
spot_wait_timeout = int(module.params.get('spot_wait_timeout'))
wait_complete = time.time() + spot_wait_timeout
spot_req_inst_ids = dict()
while time.time() < wait_complete:
reqs = ec2.get_all_spot_instance_requests()
for sirb in spot_requests:
if sirb.id in spot_req_inst_ids:
continue
for sir in reqs:
if sir.id != sirb.id:
continue # this is not our spot instance
if sir.instance_id is not None:
spot_req_inst_ids[sirb.id] = sir.instance_id
elif sir.state == 'open':
continue # still waiting, nothing to do here
elif sir.state == 'active':
continue # Instance is created already, nothing to do here
elif sir.state == 'failed':
module.fail_json(msg="Spot instance request %s failed with status %s and fault %s:%s" % (
sir.id, sir.status.code, sir.fault.code, sir.fault.message))
elif sir.state == 'cancelled':
module.fail_json(msg="Spot instance request %s was cancelled before it could be fulfilled." % sir.id)
elif sir.state == 'closed':
# instance is terminating or marked for termination
# this may be intentional on the part of the operator,
# or it may have been terminated by AWS due to capacity,
# price, or group constraints in this case, we'll fail
# the module if the reason for the state is anything
# other than termination by user. Codes are documented at
# http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/spot-bid-status.html
if sir.status.code == 'instance-terminated-by-user':
# do nothing, since the user likely did this on purpose
pass
else:
spot_msg = "Spot instance request %s was closed by AWS with the status %s and fault %s:%s"
module.fail_json(msg=spot_msg % (sir.id, sir.status.code, sir.fault.code, sir.fault.message))
if len(spot_req_inst_ids) < count:
time.sleep(5)
else:
return spot_req_inst_ids.values()
module.fail_json(msg = "wait for spot requests timeout on %s" % time.asctime())
def enforce_count(module, ec2, vpc):
exact_count = module.params.get('exact_count')
@ -1103,23 +1160,7 @@ def create_instances(module, ec2, vpc, override_count=None):
# Now we have to do the intermediate waiting
if wait:
spot_req_inst_ids = dict()
spot_wait_timeout = time.time() + spot_wait_timeout
while spot_wait_timeout > time.time():
reqs = ec2.get_all_spot_instance_requests()
for sirb in res:
if sirb.id in spot_req_inst_ids:
continue
for sir in reqs:
if sir.id == sirb.id and sir.instance_id is not None:
spot_req_inst_ids[sirb.id] = sir.instance_id
if len(spot_req_inst_ids) < count:
time.sleep(5)
else:
break
if spot_wait_timeout <= time.time():
module.fail_json(msg = "wait for spot requests timeout on %s" % time.asctime())
instids = spot_req_inst_ids.values()
instids = await_spot_requests(module, ec2, res, count)
except boto.exception.BotoServerError as e:
module.fail_json(msg = "Instance creation failed => %s: %s" % (e.error_code, e.error_message))