I know the RoR (Ruby on Rails) community relies on error/exception logging or grouping services.
Do any of you use an exception tracking or exception logging / grouping cloud service for Python apps?
Something like Trakerr.IO (
https://trakerr.io) or Airbrake.IO (
https://airbrake.io)?
In the Python world
Sentry may be the most know one,
they have also hired
Armin Ronacher which is a know person in Python.
Sentry has all there stuff as open source.
Quote:Do any of you use an exception tracking or exception logging / grouping cloud service for Python apps?
I have not have need for this yet,i did some testing a while ago with
Sentry and
Flask,
just to see how it work.
Here some overview of these tools
1,
2.
Thanks for the response.
I had taken a look at Sentry earlier and while it seemed great at first, it's primarily targeting errors. Trakerr helped me find performance issues with database issues too as you can pass in the milliseconds that a db call took along with other non-error logs. I haven't looked at Flask but will take a look.
I am actually surprised by the number of developers who do not use an error or performance tracking tool in production. Given that verbose logging is very prohibitive in terms of cost and performance, the only fallback to catching major issues in production appears to be using some real-time metrics + error capturing technology.
(Jul-11-2017, 06:22 PM)caxis Wrote: [ -> ]Thanks for the response.
I had taken a look at Sentry earlier and while it seemed great at first, it's primarily targeting errors. Trakerr helped me find performance issues with database issues too as you can pass in the milliseconds that a db call took along with other non-error logs. I haven't looked at Flask but will take a look.
I am actually surprised by the number of developers who do not use an error or performance tracking tool in production. Given that verbose logging is very prohibitive in terms of cost and performance, the only fallback to catching major issues in production appears to be using some real-time metrics + error capturing technology.
There is a good reason to that, it's not part of the specs. The people who write the specs rarely consider the maintenance/control/operations side of things. For them it's a given.