search

killy971 / hpc-coveralls

coveralls.io support for haskell code coverage with hpc
http://hackage.haskell.org/package/hpc-coveralls
BSD 3-Clause "New" or "Revised" License
60 stars 20 forks source link

Failed to read total coverage #45

Closed jkarni closed 9 years ago

jkarni commented 9 years ago

hpc-coveralls sometimes reports Failed to read total coverage, and the corresponding page on coveralls.io in fact never does display the coverage.

Beyond the logs, I myself really have no insights, unfortunately!

killy971 commented 9 years ago

Thank you for reporting this issue. I'm going to investigate what's going on. It may take me a few days until I get back to you as in unsure if I can make time today or this week end, but I will definitely have a look!

killy971 commented 9 years ago

How many times has this happened so far? (you mentioned "sometimes") From the coveralls list of servant builds, it seems it has only happened once. There is unfortunately no hint in the travis build logs. I think it may have been an issue on coveralls side, but don't have any element to prove it... I've recently improved error handling and reporting in hpc-coveralls, which I should release soon. This may help to narrow down the cause if it ever happens again.

jkarni commented 9 years ago

I believe it happened twice. Both times it seems rebuilding fixed it. Maybe the threadDelay is too short (or, better yet, it's missing retry logic)?

killy971 commented 9 years ago

The retry logic is already implemented inside the readCoverageResult function. @jkarni I will close this issue for now, but don't hesitate to reopen it if it occurs again. You can also contact me on gitter if you want!