Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
Chukwa >> mail # dev >> a more fault-tolerant collector


Copy link to this message
-
a more fault-tolerant collector
Howdy.

This is an answer to a question Bill asked me recently: can we
redesign the Collector process to behave better if the filesystem is
unavailable?

I think we can do this by backpressure. If the write fails, the
collector should return an error to the agent. And the agent should
treat it like a post failure, and retry.  Thoughts?

--Ar

--
Ari Rabkin [EMAIL PROTECTED]
UC Berkeley Computer Science Department