asterixdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ian Maxon <ima...@uci.edu>
Subject Re: sporadic buffercache related bug
Date Wed, 02 Dec 2015 22:43:26 GMT
The tool I use to track these down in the BufferCache is
confiscatedPagesOwner and pinnedPageOwner in the BufferCache. They get
used when the DEBUG flag is on. Then you can see the stack of the
thread that last called pin/confiscate. It works decently as long as
doing the debug tracing doesn't affect the timing enough to cause the
bug to fix itself ;)

On Wed, Dec 2, 2015 at 2:29 PM, abdullah alamoudi <bamousaa@gmail.com> wrote:
> Will try to reproduce it locally but so far, I have only seen it in Jenkins.
>
> Regards,
> Abdullah.
>
> Amoudi, Abdullah.
>
> On Wed, Dec 2, 2015 at 2:09 PM, Ian Maxon <imaxon@uci.edu> wrote:
>
>> Does it reproduce locally? If so I can dig into what the orphaned page
>> might belong to.
>>
>> On Wed, Dec 2, 2015 at 1:56 PM, abdullah alamoudi <bamousaa@gmail.com>
>> wrote:
>> > Hi All,
>> > Recently, I have been seeing a bug in my builds and in other committers'
>> > builds as well. The bug shows up only in the tearDown call of the
>> execution
>> > test with the following stacktrace:
>> >
>> > java.lang.IllegalStateException: null
>> >         at
>> org.apache.asterix.test.runtime.ExecutionTestUtil.validateBufferCacheState(ExecutionTestUtil.java:79)
>> >         at
>> org.apache.asterix.test.runtime.ExecutionTestUtil.tearDown(ExecutionTestUtil.java:85)
>> >         at
>> org.apache.asterix.test.runtime.ExecutionTest.tearDown(ExecutionTest.java:70)
>> >
>> >
>> > If you think that this could be related to a recent change by you, please
>> > investigate,
>> >
>> > Amoudi, Abdullah.
>>

Mime
View raw message