Ho John,

Thank you for the bugfix, everything works well now. I would recommend
making revision 443 a stable one, instead of current 1.2.5 version.

Petr

On 9 December 2011 20:51, K. John Wu <[email protected]> wrote:
> Hi, Petr,
>
> Thanks for the feedback.  I believe that I have located the source of
> the problem -- there are a number of cases I simply neglected to
> update the number of rows in the result table ;-)
>
> Please give SVN version 443 a try when you get the chance.  Let me
> know if you find anything that still needs attention.
>
> John
>
>
> On 12/9/11 2:10 AM, Thorgrin wrote:
>> Hi, John
>>
>> Thank you for the quick response. The problems I reported are now gone
>> and I'm able to aggregate much larger data thanks to the new feature!
>>
>> There is one other thing, I've noticed. When aggregating, the number
>> of different rows is determined by the first part, so if the first
>> part will return only two rows, and second would give me 4, when used
>> together I get only the two.
>>
>> Here are the commands to reproduce:
>>
>> This gives two rows:
>>   ibis -d ~/Downloads/bin/fi_hafo/00000000010[5] -q "SELECT
>> count(*),e0id4,sum(e0id2),sum(e0id1) WHERE 1=1"
>>
>> This gives four rows:
>>   ibis -d ~/Downloads/bin/fi_hafo/00000000010[6] -q "SELECT
>> count(*),e0id4,sum(e0id2),sum(e0id1) WHERE 1=1"
>>
>> This works on both parts and gives only two rows.
>>   ibis -d ~/Downloads/bin/fi_hafo/00000000010[5,6] -q "SELECT
>> count(*),e0id4,sum(e0id2),sum(e0id1) WHERE 1=1"
>>
>> Please find the test data at
>> http://www.liberouter.org/~thorgrin/fi_hafo.tar.gz
>>
>> Petr
>>
>> PS.: I track the svn in git and I have made some changes so that I can
>> easily create an rpm package from the library. Now I only have to run
>> "git svn rebase" to get new version.
>>
>> On 9 December 2011 05:42, K. John Wu <[email protected]
>> <mailto:[email protected]>> wrote:
>>
>>     Hi, Petr,
>>
>>     Thanks for the test data and the instructions to reproduce the
>>     problems.  I was able to identify two separate problems in the new
>>     functions that perform aggregation-by-parts.  A update set of files
>>     have been checked into the SVN repository at codeforge.lbl.gov
>>     <http://codeforge.lbl.gov>.
>>
>>     Please let me know if you find any additional problems.
>>
>>     John
>>
>>     PS: Just in case you need the information, you can check out the
>>     source code with the following command
>>
>>     svn checkout https://codeforge.lbl.gov/anonscm/fastbit
>>
>>     Once you have a version checked out, you can simply do 'svn update' to
>>     get the latest version in the repository.
>>
>>
>>     On 12/8/11 8:26 AM, Thorgrin wrote:
>>     > Hi John,
>>     >
>>     > thanks for the update, the new features looks great. Unfortunately
>>     > I've bumped into some troubles with the new version.
>>     >
>>     > I've downloaded and compiled the latest version from svn (revision
>>     > 441), I hope that is the latest one. There were no problem compiling
>>     > on my machine, so the previous missing declarations of some
>>     functions
>>     > in mine compiler are now OK.
>>     >
>>     > My own app was ending in segfault on aggregation queries, so I
>>     tested
>>     > with the ibis tool. Need to say that I'm running the binary ibis
>>     tool,
>>     > not the script in examples directory, copied from the machine
>>     that the
>>     > library was build on.
>>     > The data that can be used to reproduce the error are here:
>>     >
>>     > http://www.liberouter.org/~thorgrin/fi_hafo1.tar.gz
>>     <http://www.liberouter.org/%7Ethorgrin/fi_hafo1.tar.gz>
>>     >
>>     > the command is:
>>     > ibis -d ./fi_hafo1/00000000000[1,5] -q "SELECT
>>     > count(*),e0id4,sum(e0id2) WHERE 1=1"
>>     >
>>     > You will notice that this does not indicate any error but does not
>>     > produce any table line either. Previous version always prints at
>>     least
>>     > one line (more on verbose). My app ends in segfault after it
>>     finishes
>>     > the query and tries to read from the result table, so I believe
>>     it is
>>     > the same problem and ibis only reads differently.
>>     >
>>     > In attachement are two logs using -v 10, one for version 1.2.4
>>     (which
>>     > is really svn build 436 or so), one for 1.2.5. Both has
>>     extension .out.
>>     >
>>     > When I modify the query only a bit, i get segfault from the ibis
>>     tool.
>>     > The command is:
>>     > ibis -d ~/Documents/devel/data/fi_hafo1/00000000000[1,5] -q "SELECT
>>     > count(*),e0id4,sum(e0id2),sum(e0id1) WHERE 1=1"
>>     >
>>     > The valgrind output can be found in file ibis1.2.5.segfault.
>>     >
>>     > It is entirely possible that something is amis on my part, if
>>     you have
>>     > any idea what might that be, please let me know. Also if you
>>     need any
>>     > other output or information, just let me know.
>>     >
>>     > Petr Velan
>>     >
>>     >
>>     > On 8 December 2011 00:19, K. John Wu <[email protected]
>>     <mailto:[email protected]>
>>     > <mailto:[email protected] <mailto:[email protected]>>> wrote:
>>     >
>>     >     Hi, Everyone,
>>     >
>>     >     Thanks for all the input on the memory usage issues.  There is a
>>     >     updated version of FastBit source code marked as ibis1.2.5
>>     with some
>>     >     initial implementation of aggregation operations by parts.  This
>>     >     should reduce the memory usage for aggregation operations
>>     that does
>>     >     not require all the data to be in memory.
>>     >
>>     >     The new version is available at
>>     >
>>     http://codeforge.lbl.gov/project/showfiles.php?group_id=44&release_id=329
>>     
>> <http://codeforge.lbl.gov/project/showfiles.php?group_id=44&release_id=329>
>>     >
>>     
>> <http://codeforge.lbl.gov/project/showfiles.php?group_id=44&release_id=329
>>     
>> <http://codeforge.lbl.gov/project/showfiles.php?group_id=44&release_id=329>>
>>     >
>>     >     We have also included two patches from Jon Strabala
>>     >     and Robert Wong to address a couple of different
>>     >     issues.
>>     >
>>     >     This release also fixes a number issues raised by
>>     >     Alexandre Maure, Anderson Carniel, Jon Strabal,
>>     >     Andrew Olson, Teryl Taylor, Robert Wong, and
>>     >     others.
>>     >
>>     >     Please feel free to let us know if you encounter any
>>     problems with the
>>     >     new source code.
>>     >
>>     >     John
>>     >
>>     >     PS: Just in case you are wondering about which aggregations
>>     can be
>>     >     performed by parts.  The aggregation operators, count, sum,
>>     min, and
>>     >     max can be performed by parts.  The operators, avg, var, and
>>     stddev
>>     >     can be transformed into sums and counts and are currently
>>     internally
>>     >     transformed into sums and counts.  This transformation may
>>     introduce
>>     >     very slightly different results.  The computations are done with
>>     >     double-precision floating-point numbers, which have 16-digit
>>     >     precision.  The difference should be limited to the last
>>     couple of
>>     >     digits and should not be noticeable for most users.
>>      However, those
>>     >     users with very sensitive computations should know that the
>>     converted
>>     >     expression have different round-off properties from the original
>>     >     expression.  The bad operations are distinct and median,
>>     both which
>>     >     require all data to be in memory.
>>     >
>>     >     PPS: We have made some progress in supporting joins through
>>     a new
>>     >     class called quaere.  We intend to eventually rename it to
>>     >     ibis::query.
>>     >     _______________________________________________
>>     >     FastBit-users mailing list
>>     >     [email protected]
>>     <mailto:[email protected]>
>>     <mailto:[email protected]
>>     <mailto:[email protected]>>
>>     >     https://hpcrdm.lbl.gov/cgi-bin/mailman/listinfo/fastbit-users
>>     >
>>     >
>>     >
>>     >
>>     > _______________________________________________
>>     > FastBit-users mailing list
>>     > [email protected] <mailto:[email protected]>
>>     > https://hpcrdm.lbl.gov/cgi-bin/mailman/listinfo/fastbit-users
>>
>>
_______________________________________________
FastBit-users mailing list
[email protected]
https://hpcrdm.lbl.gov/cgi-bin/mailman/listinfo/fastbit-users

Reply via email to