The rate limited dir is not random - its being used in most tests as most tests 
use the std solrconfig.xml - it's hard coded but it's just using really high 
values with the idea it would not do any limiting. This was done so I could see 
it didn't have a bad affect and test that the settings matched what was in the 
config.

The settings are so high I don't think it should ever be sleeping like it is.

- Mark

On Dec 17, 2012, at 10:45 PM, Yonik Seeley <[email protected]> wrote:

> This test only brings servers up at the beginning and that's it (i.e.
> any "recovery" would be of an empty index).
> 
> It feels like we're throwing too many random elements into higher
> level tests here - introducing too many variables at once (i.e. the
> random Directory implementations).
> 
> -Yonik
> http://lucidworks.com
> 
> 
> On Mon, Dec 17, 2012 at 7:57 PM, Mark Miller <[email protected]> wrote:
>> These are the MB/s values I have - I didn't think they would kick in during 
>> our tests?
>> 
>>    <double name="maxWriteMBPerSecDefault">1000000</double>
>>    <double name="maxWriteMBPerSecFlush">2000000</double>
>>    <double name="maxWriteMBPerSecMerge">3000000</double>
>>    <double name="maxWriteMBPerSecRead">4000000</double>
>> 
>> Mark
>> 
>> On Dec 17, 2012, at 7:35 PM, Mark Miller <[email protected]> wrote:
>> 
>>> 
>>> On Dec 17, 2012, at 7:31 PM, Mark Miller <[email protected]> wrote:
>>> 
>>>> Socket read timeouts. I'll raise it for this case.
>>> 
>>> Nevermind - it's already at 60 seconds. What's odd is that the commit is 
>>> taking that long.
>>> 
>>> One of the threads is waiting in the rate limiter - which is weird - I 
>>> thought I put it absurdly high for tests. I'll dig into it when I get a 
>>> chance:
>>> 
>>> [junit4:junit4]   1> Thread[qtp29204452-1358,5,TGRP-ShardRoutingTest]
>>> [junit4:junit4]   1> java.lang.Thread.sleep(Native Method)
>>> [junit4:junit4]   1> java.lang.Thread.sleep(Thread.java:338)
>>> [junit4:junit4]   1> 
>>> org.apache.lucene.store.RateLimiter$SimpleRateLimiter.pause(RateLimiter.java:111)
>>> [junit4:junit4]   1> 
>>> org.apache.lucene.store.RateLimitedIndexOutput.flushBuffer(RateLimitedIndexOutput.java:45)
>>> [junit4:junit4]   1> 
>>> org.apache.lucene.store.BufferedIndexOutput.flushBuffer(BufferedIndexOutput.java:90)
>>> [junit4:junit4]   1> 
>>> org.apache.lucene.store.BufferedIndexOutput.flush(BufferedIndexOutput.java:79)
>>> [junit4:junit4]   1> 
>>> org.apache.lucene.store.RateLimitedIndexOutput.flush(RateLimitedIndexOutput.java:69)
>>> [junit4:junit4]   1> 
>>> org.apache.lucene.store.BufferedIndexOutput.close(BufferedIndexOutput.java:103)
>>> [junit4:junit4]   1> 
>>> org.apache.lucene.store.RateLimitedIndexOutput.close(RateLimitedIndexOutput.java:78)
>>> [junit4:junit4]   1> 
>>> org.apache.lucene.store.ChecksumIndexOutput.close(ChecksumIndexOutput.java:61)
>>> [junit4:junit4]   1> 
>>> org.apache.lucene.index.SegmentInfos.finishCommit(SegmentInfos.java:927)
>>> [junit4:junit4]   1> 
>>> org.apache.lucene.index.IndexWriter.finishCommit(IndexWriter.java:2826)
>>> [junit4:junit4]   1> 
>>> org.apache.lucene.index.IndexWriter.commitInternal(IndexWriter.java:2815)
>>> [junit4:junit4]   1> 
>>> org.apache.lucene.index.IndexWriter.commit(IndexWriter.java:2788)
>> 
>> 
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [email protected]
>> For additional commands, e-mail: [email protected]
>> 
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
> 


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to