Hi Sean,
That helps to know. Might help if the wiki talks about what the
timeout value doesn't control.
-J
On Thu, Mar 3, 2011 at 8:08 AM, Sean Cribbs wrote:
> Yes, it's possible that an individual phase can timeout before the entire job
> has timed out. We realize this isn't intuitive and ar
Yes, it's possible that an individual phase can timeout before the entire job
has timed out. We realize this isn't intuitive and are looking into solutions.
Sean Cribbs
Developer Advocate
Basho Technologies, Inc.
http://basho.com/
On Mar 3, 2011, at 9:44 AM, Jason J. W. Williams wrote:
> Hi S
Hi Sean,
That's weird, because I'm seeing timeouts sooner than the one passed
in but only when it's the reduce phase that errors out.
-J
On Thu, Mar 3, 2011 at 6:34 AM, Sean Cribbs wrote:
> Jason,
>
> You can't currently change the timeout on individual phases, just on the
> entire job. We're
Jason,
You can't currently change the timeout on individual phases, just on the entire
job. We're reprioritizing MapReduce to receive some extra love in the near
future, so if that is a critical feature for you, be sure to file a bug on
issues.basho.com.
Sean Cribbs
Developer Advocate
Basho T
Hi Sean,
Thanks. Also, is there anyway to increase the javascript reduce phase
timeout? Doing some test M/R jobs, and the reduce phase doesn't seem
to honor the timeout passed in? Thought playing with streaming would
help that, but looks like not. Thank you in advance.
-J
On Wed, Mar 2, 2011 at
If your query has "keep":true on multiple phases, that phase key lets you
disambiguate results.
Sean Cribbs
On Mar 2, 2011, at 9:28 PM, "Jason J. W. Williams"
wrote:
> Hello,
>
> When running map/reduce in streaming mode, since the results are from
> the last phase (reduce) what is the sign
