Failure to perform action due to tcp i/o timeout

[Christopher Lee]

Actions (e.g. block, attach, deploy) are failing with the error 'read tcp <ip address> i/o timeout.

As seen in this issue report: http://reports.vapour.ws/releases/issue/57855931749a564ea5a83069

The error looks similar to:
  ERROR cmd supercommand.go:458 read tcp 127.0.0.1:50948->127.0.0.1:37017: i/o timeout
or:
  ERROR cmd supercommand.go:458 failed to upload resource "bar": PUT https://10.223.16.64:17070/model/e49c27e1-9c0c-4e18-83f1-4fa761b982ff/applications/dummy-resource/resources/bar: read tcp 127.0.0.1:41528->127.0.0.1:37017: i/o timeout

An example failure log can be found here: http://reports.vapour.ws/releases/4132/job/functional-resources/attempt/156

[Ian Booth (wallyworld)]

These errors are typically caused when the system is busy and i/o bound. And as seen, various operations exhibit the symptom. Once solution would be to make use of juju/retry in more places in the code.

[Anastasia (anastasia-macmood)]

May have been fixed as part of other improvements - last failure was on Oct 10.

[Launchpad Janitor (janitor)]

[Expired for juju because there has been no activity for 60 days.]