Closed Bug 1090412 Opened 11 years ago Closed 11 years ago

Move Tokenserver+Verifier Stage to new development IAM

Categories

(Cloud Services :: Operations: Deployment Requests - DEPRECATED, task, P1)

task

Tracking

(Not tracked)

VERIFIED FIXED

People

(Reporter: jbonacci, Unassigned)

Details

(Whiteboard: [qa+])

Because we need an official bug for QA to track.
token.stage.mozaws.net points at the new stack. This also has a new RDS server so we might need to do some work if we need the data from the old RDS server.
Status: NEW → RESOLVED
Closed: 11 years ago
Resolution: --- → FIXED
I can't imagine why we would. No active testing is happening in Stage that would require transfer of data... Anyway, cool. I will start verifying this asap...
Still in the process of checking things out, but just pointing out a few changes: Instance size is m3.medium here. I think the last deploy to old Dev IAM it was a c3.large. There are fewer "token-prod.ini" processes running: 4 instead of 6 I could not find this token process: token local/bin/python -m tokenserver.scripts.process_account_deletions -v ./token-prod.ini ts-account-change-stage The verifier_out.log shows a couple of these from today: {"op":"bid.v2","name":"bid.v2","time":"2014-10-29T20:07:59.786Z","pid":3015,"v":1,"hostname":"ip-172-31-39-215","message":"verify","result":"failure","reason":"timeout trying to load well-known for api-accounts.stage.mozaws.net","assertion":"...etc...,"trustedIssuers":["api.accounts.firefox.com","api-accounts.stage.mozaws.net","api-accounts-dev.stage.mozaws.net","api-accounts.dev.lcip.org","api-accounts-latest.dev.lcip.org","nightly.dev.lcip.org","latest.dev.lcip.org"],"rp":"https://token.stage.mozaws.net"} The process_account_deletions.error.log shows a lot of traceback from today and earlier: Traceback (most recent call last): File "/usr/lib64/python2.6/runpy.py", line 122, in _run_module_as_main "__main__", fname, loader, pkg_name) File "/usr/lib64/python2.6/runpy.py", line 34, in _run_code exec code in run_globals File "/data/tokenserver/tokenserver/scripts/process_account_deletions.py", line 121, in <module> tokenserver.scripts.run_script(main) File "tokenserver/scripts/__init__.py", line 19, in run_script exitcode = main() File "/data/tokenserver/tokenserver/scripts/process_account_deletions.py", line 116, in main opts.aws_region, opts.queue_wait_time) File "/data/tokenserver/tokenserver/scripts/process_account_deletions.py", line 81, in process_account_deletions queue.delete_message(msg) File "/data/tokenserver/local/lib/python2.6/site-packages/boto/sqs/queue.py", line 299, in delete_message return self.connection.delete_message(self, message) File "/data/tokenserver/local/lib/python2.6/site-packages/boto/sqs/connection.py", line 208, in delete_message params = {'ReceiptHandle' : message.receipt_handle} AttributeError: 'str' object has no attribute 'receipt_handle' Other than that, everything checks out ok. I can load test once the loads cluster is back up and running (it is being migrated right now I think)... :mostlygeek please check the logs and let me know if that is all expected...
Flags: needinfo?(bwong)
Hmm... seems like it is a bug in boto. Tokenserver lists boto=2.27, boto=2.34 is out. Maybe upgrading to that will fix this issue. Also we're on tokenserver 1.2.9, and there is a 1.2.10 with what looks like some pretty extensive changes. Maybe should roll 1.2.11 with the more recent boto and give it a test. :rfkelly what do you think?
Flags: needinfo?(bwong) → needinfo?(rfkelly)
> Maybe should roll 1.2.11 with the more recent boto and give it a test. Yes, we should definitely do this. I'll prep it today.
Flags: needinfo?(rfkelly)
OK, so POC that the migration was a happy one. Next deployment ticket for TS+Verifier will go to new Dev IAM.
Status: RESOLVED → VERIFIED
You need to log in before you can comment on or make changes to this bug.