[Orca-users] SE Toolkit 3.4 and Solaris 10

Liston Bias bias at pobox.com
Mon Feb 14 08:29:22 PST 2005


Following version of perl used:

   ActivePerl-5.8.6.811-sun4-solaris


On Sun, 13 Feb 2005, Blair Zajac wrote:

> Liston Bias wrote:
>> 
>> I installed SE Toolkit 3.4 and latest orca snapshot on Sunfire 240 running 
>> Solaris 10.  I ran into some issues, but it seems to be working well now 
>> although I haven't tried verifying the data in the graph yet.
>
> Hi Liston,
>
> Welcome back!  Haven't heard from you lately.  Thanks for testing the latest 
> build.
>
>> Snapshot used was r412 from:
>> 
>>   http://www.orcaware.com/orca/pub/snapshots/
>> 
>> 
>> 1)  When I installed, I get a file for lib/SE/3.4 instead of directory (the 
>> file looks like workinfo_class.se).
>> 
>> This Error prevented the orcallator data gathering process from starting.
>> 
>> I seem to have fixed this by changing 3.4 to a directory, and populating 
>> with stuff from:
>> 
>>   http://svn.orcaware.com/repos/trunk/orca/lib/SE/3.4/
>
> I patched the lib/Makefile.in to create the 3.4 directory in the install 
> tree, so this should work now.
>
>> 2)  The Graphs for following showed up as invalid links.
>> 
>> Daily alpha Interface Deferred Packet Rate
>> Daily alpha Interface Collisions
>> Daily alpha Interface Nocanput Rate
>> Daily alpha Disk Run Percent
>> Daily alpha Disk Service Time
>> 
>> The explanation and patch indicated in prior posting to list seems to 
>> resolve this issue:
>> 
>>   http://www.orcaware.com/pipermail/orca-dev/2004-July/000528.html
>
> I'll look into putting this patch into Orca.
>
>> 3)  When I start the orca graph generation, I was the following error:
>> 
>> Argument "2.121_02" isn't numeric in subroutine entry at 
>> /usr/local/orca/bin/orca line 51.
>> 
>> The line listed is where the data dumper module is specified, so I reloaded 
>> it with following... And, even though, it failed half fo the make test it 
>> seems to have solved the issues and I see no other errors.
>
> I just saw this the other day with a fresh install of ActiveState's 
> ActivePerl Build 811.  I didn't try reinstalling Data::Dumper.  This is an 
> odd bug, the first time.
>
> What version of Perl are you using?  Who built it?
>
>> 
>>       % wget 
>> http://www.perl.com/CPAN/authors/id/G/GS/GSAR/Data-Dumper-2.101.tar.gz
>>       % gunzip -c Data-Dumper-2.101.tar.gz | tar xvf -
>>       % cd Data-Dumper-2.101
>>       % perl Makefile.PL
>>       % make
>>       % make test
>>       % make install UNINST=1
>> 
>> *** From make test
>>         Failed 78/162 tests, 51.85% okay
>> Failed 1/2 test scripts, 50.00% okay. 78/163 subtests failed, 52.15% okay.
>> make: *** [test_dynamic] Error 29
>
> This isn't good.  I just tried rebuilding Data::Dumper with ActiveState 
> Perl's and it passed just fine.
>
>> 4) Lastly, The Sunfire 240 has bge interfaces for which I am seeing traffic 
>> on the graphs okay.  It also showing me output graphs for ge interfaces for 
>> which I have NONE.  This really doesn't concern me, but was wondering if 
>> anyone else has similar similar behavior.
>
> Which interfaces?  Can you send the first line of your output orcallator data 
> files?
>
> Thanks,
> Blair
>
> -- 
> Blair Zajac <blair at orcaware.com>
> Plots of your system's performance - http://www.orcaware.com/orca/
>




More information about the Orca-users mailing list