[collectd] Issue #430 -- write_graphite fails after carbon-cache.py is restarted

[Michael Hart]

In collectd 5.4, there is a bug in write_graphite.c in wg_send_buffer() which I think was introduced with the “LogSendErrors” parameter. The workaround in 5.4 is to always set LogSendErrors to true (the default) and then collectd will always reconnect. At least... it always does in my simple tests. In offline conversations with Andreas, this issue appears in 5.1, but I have no explanation for why it occurs there.

I will submit a pull request shortly for a fix.

cheers
mike

--
Michael Hart
Arctic Wolf Networks
M: 226-388-4773





On Dec 19, 2013, at 6:31 AM, Andreas Rütten <andreas at smaato.com<mailto:andreas at smaato.com>> wrote:

Hi,

I just want to share that we are also seeing the effect that all our
collectd clients stop sending data to carbon after we restart it. After
we restart also the collectd clients everything works as expected again.


Is there somebody working on this issue?
Or is there some missing information which would help?


Cheers,
Andreas


PS:
Please CC me as I'm not on the list

--
Andreas Rütten
Senior System Administrator

Smaato Inc.
San Francisco - Hamburg - Singapore
www.smaato.com<http://www.smaato.com>

Germany:
Gerhofstrasse 2
20354 Hamburg
T +49 (0)40 3480 9490
F +49 (0)40 4921 9055

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.verplant.org/pipermail/collectd/attachments/20131220/959be313/attachment.html>