Our legacy Linux infrastructure provider is performing maintenance on some host machines in order to reduce error rates. Job restart counts are declining as a result.
Posted May 27, 2015 - 16:37 UTC
Identified
We have found a high number of API errors on our legacy Linux provider to be the cause of job restarts, and are working with them on resolution.
Posted May 27, 2015 - 16:07 UTC
Investigating
We are investigating reports of high numbers of build restarts on both public and private repositories. A recent spike in timeout errors for both Redis and Postgres may be related.