wikimedia / PyBal / 210 / 1

74%

master: 97%

push

travis-ci

<a href="https://github.com/wikimedia/PyBal/commit/<a class=hub.com/wikimedia/PyBal/commit/<a class="double-link" href="https://git"><a class=hub.com/wikimedia/PyBal/commit/2677056b1672be00e8aea23408268909a27e0c7c">2677056b1<a href="https://github.com/wikimedia/PyBal/commit/2677056b1672be00e8aea23408268909a27e0c7c">&lt;a href=&quot;https://github.com/wikimedia/PyBal/commit/2677056b1672be00e8aea23408268909a27e0c7c&quot;&gt;&amp;quot;&amp;gt;etcd: Introduce reconnectTimeout

In production, we have observed an issue with pybal losing its
connections to etcd shortly after a bunch of unhandled
twisted.internet.error.ConnectionLost errors.

See https://phabricator.wikimedia.org/P6711 for an example of such
situation.

In a test environment, we have verified that pybal does not reconnect to
etcd in case of etcd restarts.

Introduce a 1 second timeout between reconnection attempts to be used in
case of lost connections or connection failures and log the reconnection
attempts themselves. This fixes the reconnection issue in a testing
environment when issuing a systemctl etcd restart as well as when
stopping etcd altogher and starting it again after a few seconds.

Bug: T169765
Change-Id: I&lt;/a&gt;&lt;a class=&quot;double-link&quot; href=&quot;https://github.com/wikimedia/PyBal/commit/&lt;a class=&quot;double-link&quot; href=&quot;https://github.com/wikimedia/PyBal/commit/0db3a64af405aa11aaf3a215a8d57a8857d48dbd">0db3a64af</a>">0db3a64af</a><a href="https://github.com/wikimedia/PyBal/commit/2677056b1672be00e8aea23408268909a27e0c7c">
(cherry picked from commit </a><a class="double-link" href="https://github.com/wikimedia/PyBal/commit/<a class="double-link" href="https://github.com/wikimedia/PyBal/commit/c917f59f4c37c7b89f15d0fd8b375de028d4a90b">c917f59f4">c917f59f4)

2862 of 3872 relevant lines covered (73.92%)

0.74 hits per line