Hi, guys:

 I follow your rangesever crashes problem several days. I am a freshman on
Hypertable and interested on it. Please allow me to say something :-)

 SigAbort means short of memory?  so we can check if it is really the
sistuation.
 if not, the problem may be not simple.  We maybe face the memory fragment
problem. Maybe it is not suitable just to use new to malloc a big chunck of
memory?


2012/8/15 <[email protected]>

>   Today's Topic Summary
>
> Group: http://groups.google.com/group/hypertable-dev/topics
>
>    - RangeServer crashes <#13929fdab39e9bab_group_thread_0> [3 Updates]
>    - RangeServer/DfsBroker failled to come up after 
> restart<#13929fdab39e9bab_group_thread_1>[1 Update]
>
>   RangeServer 
> crashes<http://groups.google.com/group/hypertable-dev/t/5f5af3bea0d3046b>
>
>    "Kenny F." <[email protected]> Aug 14 06:51AM -0700
>
>    0.9.6.0.95b0abc tracks:
>
>    # screen -r
>
>    Program received signal SIGABRT, Aborted.
>    [Switching to Thread 0x99ec9b70 (LWP 15640)]
>    0xffffe424 in __kernel_vsyscall ()
>
>    # where
>
>    #0 0xffffe424 in __kernel_vsyscall ()
>    #1 0xb7b28781 in raise () from /lib/i686/cmov/libc.so.6
>    #2 0xb7b2bbb2 in abort () from /lib/i686/cmov/libc.so.6
>    #3 0xb7d34959 in __gnu_cxx::__verbose_terminate_handler() () from
>    /opt/hypertable/0.9.6.0.95b0abc/lib/libstdc++.so.6
>    #4 0xb7d32865 in ?? () from
>    /opt/hypertable/0.9.6.0.95b0abc/lib/libstdc++.so.6
>    #5 0xb7d328a2 in std::terminate() () from
>    /opt/hypertable/0.9.6.0.95b0abc/lib/libstdc++.so.6
>    #6 0xb7d329da in __cxa_throw () from
>    /opt/hypertable/0.9.6.0.95b0abc/lib/libstdc++.so.6
>    *#7 0xb7d33033 in operator new(unsigned int) () from
>    /opt/hypertable/0.9.6.0.95b0abc/lib/libstdc++.so.6
>    #8 0xb7d3311d in operator new[](unsigned int) () from
>    /opt/hypertable/0.9.6.0.95b0abc/lib/libstdc++.so.6*
>    #9 0x0869ef81 in Hypertable::DynamicBuffer::grow (this=0x99ec7ff4,
>    new_size=1000004, nocopy=false) at
>    /root/src/hypertable/src/cc/Common/DynamicBuffer.h:120
>    #10 0x0869f095 in Hypertable::DynamicBuffer::reserve (this=0x99ec7ff4,
>    len=1000004, nocopy=false) at
>    /root/src/hypertable/src/cc/Common/DynamicBuffer.h:72
>    #11 0x08773887 in Hypertable::FillScanBlock (scanner=..., dbuf=...,
>    buffer_size=1000000) at
>    /root/src/hypertable/src/cc/Hypertable/RangeServer/FillScanBlock.cc:104
>    #12 0x086628c1 in Hypertable::RangeServer::create_scanner
>    (this=0x8c95460,
>    cb=0x99ec92a8, table=0x99ec929c, range_spec=0x99ec9290,
>    scan_spec=0x99ec9200, cache_key=0x99ec9278)
>    at
>    /root/src/hypertable/src/cc/Hypertable/RangeServer/RangeServer.cc:1371
>    #13 0x087d53e6 in Hypertable::RequestHandlerCreateScanner::run
>    (this=0x331d9840) at
>
>    
> /root/src/hypertable/src/cc/Hypertable/RangeServer/RequestHandlerCreateScanner.cc:59
>    #14 0x0862debc in Hypertable::ApplicationQueue::Worker::operator()
>    (this=0x8c83cf8) at
>    /root/src/hypertable/src/cc/AsyncComm/ApplicationQueue.h:172
>    #15 0x0862df18 in
>    boost::detail::thread_data<Hypertable::ApplicationQueue::Worker>::run
>    (this=0x8c83c28) at
>    /usr/local/include/boost/thread/detail/thread.hpp:61
>    #16 0xb7ebfe68 in thread_proxy () from
>    /opt/hypertable/0.9.6.0.95b0abc/lib/libboost_thread.so.1.44.0
>    #17 0xb7e05955 in start_thread () from /lib/i686/cmov/libpthread.so.0
>    #18 0xb7bca5ee in clone () from /lib/i686/cmov/libc.so.6
>
>
>
>
>    Mehmet Ali Cetinkaya <[email protected]> Aug 14 07:38AM -0700
>
>    Hello,
>
>    i have a Hadoop (0.20)+Hypertable (0.9.6) system that one master and
>    two slaves.
>
>    we inserted 1 million data to hypertable succesfully.
>    But now, when use the select query ( select meta:eklemetarihi from
>    urls; )
>    hypertable frozen after approximate 300.000 data show.
>
>    i didn't see any error in hypertable logs.
>
>    i read some document from internet. and their solution is "you must
>    erase cell cache of hypertable".
>    but i didn't find how can i erase cell cache.
>
>    what can be done to resolvee this issue?
>
>    thanx,
>    mali
>
>
>
>
>    Doug Judd <[email protected]> Aug 14 03:53PM -0700
>
>    Hi Mali,
>
>    This could be a number of things. First, check to make sure the
>    RangeServers (slaves) are idle when the query appears to be hanging.
>    I've
>    witnessed situations where a query appears to hang, but is actually
>    still
>    being executed, scanning over a large section of data that, for
>    example,
>    does not contain the column "meta:eklemetarihi". The next thing to do
>    is
>    to double-check that there are no errors in the Hypertable logs. You
>    can
>    do this with:
>
>    grep -i ERROR /opt/hypertable/current/log/*.log
>    grep -i Except /opt/hypertable/current/log/*.log
>
>    Next double-check the Hadoop logs to be sure there are no errors (see
>    your
>    Hadoop distro documentation for location of logs). If none of the above
>    turns up anything, let us know and we can help you capture stack
>    traces of
>    the RangeServers to determine if they're stuck in a deadlock.
>
>    - Doug
>
>    On Tue, Aug 14, 2012 at 7:38 AM, Mehmet Ali Cetinkaya <
>
>    --
>    Doug Judd
>    CEO, Hypertable Inc.
>
>
>
>   RangeServer/DfsBroker failled to come up after 
> restart<http://groups.google.com/group/hypertable-dev/t/ab5cdbc8fdd3329e>
>
>    herb <[email protected]> Aug 14 09:10AM -0700
>
>    Christoph-
>
>    How would one go about deleting a range that landed in this data loss?
>    Ideally we could do this wholesale as we run into these CellStores.
>
>
>
>
>  You received this message because you are subscribed to the Google Group
> hypertable-dev.
> You can post via email <[email protected]>.
> To unsubscribe from this group, 
> send<[email protected]>an empty message.
> For more options, 
> visit<http://groups.google.com/group/hypertable-dev/topics>this group.
>
> --
> You received this message because you are subscribed to the Google Groups
> "Hypertable Development" group.
> To post to this group, send email to [email protected].
> To unsubscribe from this group, send email to
> [email protected].
> For more options, visit this group at
> http://groups.google.com/group/hypertable-dev?hl=en.
>

-- 
You received this message because you are subscribed to the Google Groups 
"Hypertable Development" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/hypertable-dev?hl=en.

Reply via email to