hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sudarshan Kadambi (BLOOMBERG/ 731 LEXIN)" <skada...@bloomberg.net>
Subject Re: Stack overflow errors
Date Tue, 25 Feb 2014 19:34:50 GMT
Ted:

To your previous question about HA mode - on the test cluster where this is, we have a single
node HA - there's only one namenode specified in dfs.ha.namenodes.HDFSNAME. Do you think this
configuration is the root of this error?

----- Original Message -----
From: user@hbase.apache.org
To: Sudarshan Kadambi (BLOOMBERG/ 731 LEXIN)
Cc: user@hbase.apache.org
At: Feb 25 2014 12:04:51

Looking at the stack trace for IllegalArgumentException, it reminds me of
HBASE-10029: Proxy created by HFileSystem#createReorderingProxy() should
properly close when connecting to HA namenode

The above issue has been resolved in the recently released 0.98.0

FYI


On Tue, Feb 25, 2014 at 7:59 AM, Ted Yu <yuzhihong@gmail.com> wrote:

> In the jstack pastebin, I don't see StackOverflowError.
> Not in region server log, either.
>
> I saw the following:
>
> java.lang.IllegalArgumentException: object is not an instance of declaring class
>
> Are you using hdfs HA ?
>
>
>
> On Tue, Feb 25, 2014 at 7:54 AM, Sudarshan Kadambi (BLOOMBERG/ 731 LEXIN)
> <skadambi@bloomberg.net> wrote:
>
>> Ted - thanks for taking a look.
>>
>> regionserver.log -> http://paste.ubuntu.com/6995011/
>> regionserver.out -> http://paste.ubuntu.com/6994999/
>> jstack -> http://paste.ubuntu.com/6995016/
>>
>> ----- Original Message -----
>> From: yuzhihong@gmail.com
>> To: Sudarshan Kadambi (BLOOMBERG/ 731 LEXIN) <skadambi@bloomberg.net>,
>> user@hbase.apache.org
>> At: Feb 25 2014 10:41:48
>>
>> Do you have more of the stack trace ?
>> If so, please pastebin it.
>>
>> Thanks
>>
>>
>> On Tue, Feb 25, 2014 at 7:21 AM, Sudarshan Kadambi (BLOOMBERG/ 731 LEXIN)
>> <skadambi@bloomberg.net> wrote:
>>
>>> I recently saw some exceptions that I haven't seen in HBase before. The
>>> trace looks something like:
>>>
>>> Exception in thread "RpcServer.handler=96,port=60020"
>>> java.lang.StackOverflowError
>>>         at org.apache.hadoop.hbase.CellUtil$1.advance(CellUtil.java:160)
>>>
>>> Is this a known issue? The RegionServer went into a hung state and all
>>> gets to that RS began to hang indefinitely. This is on HBase 0.96.
>>
>>
>>
>

Mime
View raw message