1326204 - Give Frank Access to Papertrail

Status: RESOLVED WONTFIX

(Cloud Services Graveyard :: Metrics: Pipeline, defect)

Description

[Frank Bertsch [:frank]]

We're going to use it for the aggregates endpoint.

Comment 1

[Thomas Huelbert]

confusion generated by bug :) - uncertain as to what you are trying achieve. Can you be more specific?

Comment 2

[Frank Bertsch [:frank]]

Whoops! I went ahead and just created my own account. I may have to eventually integrate with some sort of company payment if the usage gets high enough.

Comment 3

[Roberto Agostino Vitillo (:rvitillo)]

(In reply to Frank Bertsch [:frank] from comment #2)
> Whoops! I went ahead and just created my own account. I may have to
> eventually integrate with some sort of company payment if the usage gets
> high enough.

Please, let's not use personal accounts. Jason, do we have a Papertrail account for Mozilla?

Comment 4

[Jason Thomas [:jason]]

As far as I know Mozilla doesn't have a global papertrail account, I believe it is team/group specific. Cloud services operations currently doesn't use papertrail. I think releng/releng ops may have a account. 

Do you feel like we should use that account instead here? Maybe we can deploy a solution without using Papertrail?

Comment 5

[Frank Bertsch [:frank]]

I had a simple enough solution using [0], which would bypass paper trail, but leave more work for viewing requests and obviously more code.

I know that ATMO uses papertrail, maybe Mauro can tell us which account they're using.

[0] https://github.com/fbertsch/s3_stream_logger

Comment 6

[Mauro Doglio [:mdoglio]]

:jason does Cloud services operations have something similar to papertrail in house or as a 3rd party service? 
atmo was using papertrail back in the heroku days, and that was automatically added to the heroku bill.
I know several projects run by Platform Operations (e.g. Taskcluster, Treeherder) use Papertrail a lot, their account name is "Mozilla Taskcluster".

Comment 7

[Jason Thomas [:jason]]

Cloud Services Operations have what we call 'Logging 2.0' which is used for our Cloud Services managed services [1]. We could potentially integrate it with aggregates.t.m.o but the level of effort would be high since we won't be able to use the automation pieces completely.

AWS Cloudwatch could also be a option especially if the volume of logs is low [2].

We currently do not have any plans to support Papertrail. It would be good to find out what the logging requirements are and see if the above solutions make sense.


[1] https://mana.mozilla.org/wiki/display/SVCOPS/Logging+2.0 
[2] http://docs.aws.amazon.com/AmazonCloudWatch/latest/logs/WhatIsCloudWatchLogs.html

Comment 8

[Jason Thomas [:jason]]

Please let me know if you still need a solution and we can discuss further.