headers
pete wright | 1 Dec 2005 01:01
Picon

Re: Oracle 10g on FreeBSD 6.0

On 11/30/05, Anthony Elizondo <anthony.elizondo <at> gmail.com> wrote:
> On 11/30/05, Steve Rieger <steve.rieger <at> tbwachiat.com> wrote:
> >
> > you need to install linux compat,
> >
> > i once did it on oracle 9, and never again
>
> Does "linux compat" mean linux_base?
>
> I forgot to mention: From ports I installed java/linux-sun-jdk15,
> devel/linux_devtools-8, and emulators/linux_base-8, which gave me the
> full complement of stuff in /compat/linux/

if you have the linux.ko loaded you should be all set.  i'd check to
make sure you have linprocfs loaded as well, as i believe the oracle
installer utilizies it.

and yea, oracle+linux=no fun okan :)

-p

--
~~o0OO0o~~
Pete Wright
www.nycbug.org
NYC's *BSD User Group
Permalink | Reply | 8 comments |
headers
Anthony Elizondo | 1 Dec 2005 16:18
Picon

Re: Oracle 10g on FreeBSD 6.0

On 11/30/05, pete wright <nomadlogic <at> gmail.com> wrote:
> if you have the linux.ko loaded you should be all set.  i'd check to
> make sure you have linprocfs loaded as well, as i believe the oracle
> installer utilizies it.
>
> and yea, oracle+linux=no fun okan :)

I have linux.ko loaded.

Hmm, I'm stuck. From within the bash shell at /compat/linux/bin/bash I
have awk, sed, and the seperate chown, so I don't think I need to make
those symlinks. I made /compat/linux/bin/arch and it returns "i686".
How does the installer determine the operating system...

Wait, found it. /compat/linux/etc/redhat-release currently holds only one line:

Red Hat Linux release 8.0 (Psyche)

I've added the following line, stolen from Red Hat EL 4:

Red Hat Enterprise Linux AS release 4 (Nahant)

Onward!
Permalink | Reply | 7 comments |
headers
Anthony Elizondo | 1 Dec 2005 16:21
Picon

Re: Oracle 10g on FreeBSD 6.0

I forgot to mention: I'm doing this partly for educational enrichment.
But I am also doing it to show people in my company that FreeBSD can
do what Linux can, sometimes better. I'll be doing some simple
performance testing to see if Oracle+FreeBSD can compete with
Oracle+Linux.

We are an Oracle shop. Sorry, but that part of the equation is, at
this point, not negotiable. :)
Permalink | Reply | 6 comments |
headers
pete wright | 1 Dec 2005 17:27
Picon

Re: Oracle 10g on FreeBSD 6.0

On 12/1/05, Anthony Elizondo <anthony.elizondo <at> gmail.com> wrote:
> I forgot to mention: I'm doing this partly for educational enrichment.
> But I am also doing it to show people in my company that FreeBSD can
> do what Linux can, sometimes better. I'll be doing some simple
> performance testing to see if Oracle+FreeBSD can compete with
> Oracle+Linux.
>

well if you are going to do a performance test between two platforms
atleast use a common ground to do the test.  Oracle is seriously tuned
for their supported platforms.  I don't even think they support
RHEL4.x at this point.  PostgreSQL (heh I guess you can tell where
most of stand in the DB wars :) may be a good option for Database
performace benchmarking (MySQL obviously would be another good
application as well).  They are both supported and  well documented
under gnu/linux and *BSD.  Now I'm not saying that *BSD is not able to
run Oracle, but I would read any results gotten from this test with a
grain of salt while trying to compare the functionality and performace
of two platforms.

> We are an Oracle shop. Sorry, but that part of the equation is, at
> this point, not negotiable. :)

Yea, one can't argue with that :)  While you are doing benchmarking it
may be fun to see how oracle behaves on linux and solaris.  Now I
think that would be a more fair comparison.

-pete

--
(Continue reading)

Permalink | Reply | 0 comments |
headers
Bjorn Nelson | 1 Dec 2005 19:49

multiple make buildkernel

NYCBuggers,

I have a quad processor box that is fairly slow cpu-wise.  I am using  
the make buildworld -j option (benchmarks so far are 47 min 9 sec at  
j11, 47 min 18 sec at j10,  47 min 5 sec at j9, still playing with  
this).  Buildkernel seems to take an hour and half at least.  I am  
wondering what would be involved to get multiple buildkernel's to run  
for different kernels.  I believe a make buildkernel only writes to / 
usr/obj/usr/src/sys/KERNELNAME.  If this is true, it should be okay  
for multiple buildkernel's to run, right?

Thanks,
Bjorn Nelson
Permalink | Reply | 2 comments |
headers
N.J. Thomas | 1 Dec 2005 20:39

overloaded webserver: nfs wait issue?

We have a website with moderately high traffic, load balanced among 3
webservers.

During peak traffic times however (when the volume is higher than
normal), the load shoots up to over a 100, and the site crawls to its
knees.

We set up a script to take snapshots of top every 20 seconds. Here is
what it looks like when everthing is normal:

         127
    last pid: 12003;  load averages:  0.93,  1.36,  1.35  up 41+04:22:14    14:00:23
    243 processes: 12 running, 230 sleeping, 1 zombie

    Mem: 222M Active, 74M Inact, 186M Wired, 16M Cache, 111M Buf, 503M Free
    Swap: 2048M Total, 16M Used, 2032M Free

      PID USERNAME     PRI NICE  SIZE    RES STATE    TIME   WCPU    CPU COMMAND
      136 root          32   0  1208K   420K RUN     33.1H  7.28%  7.28% amd
    11918 nobody        -1   0   149M 12292K nfsrcv   0:01  3.00%  1.95% httpd
    11879 nobody         2   0   149M 12292K sbwait   0:01  2.10%  1.37% httpd
    11896 nobody         2   0   148M 11704K RUN      0:00  1.80%  1.17% httpd
    11962 nobody         2   0   147M 10072K RUN      0:00  4.33%  1.12% httpd
    11892 nobody        -1   0   145M  8804K nfsrcv   0:00  1.35%  0.88% httpd
    11935 nobody         2   0   149M 12284K sbwait   0:00  1.73%  0.78% httpd
    11925 nobody         2   0   149M 12288K sbwait   0:00  1.08%  0.68% httpd
    11894 nobody         2   0   149M 12404K sbwait   0:00  0.98%  0.63% httpd
    11937 nobody         2   0   149M 12456K RUN      0:00  1.61%  0.63% httpd
    11954 nobody         2   0   149M 12288K sbwait   0:00  1.88%  0.49% httpd
      191 root           2   0   144M  6632K select  13:23  0.34%  0.34% httpd
(Continue reading)

Permalink | Reply | 4 comments |
headers
N.J. Thomas | 1 Dec 2005 21:08

Re: overloaded webserver: nfs wait issue?

* N.J. Thomas <njt <at> ayvali.org> [2005-12-01 14:39:44 -0500]:
> We have a website with moderately high traffic, load balanced among 3
> webservers.
> 
> During peak traffic times however (when the volume is higher than
> normal), the load shoots up to over a 100, and the site crawls to its
> knees.

I forgot to mention that the webservers are running FreeBSD 4.11 and
Apache 1.3.x.

Thomas

--

-- 
N.J. Thomas
njt <at> ayvali.org
Etiamsi occiderit me, in ipso sperabo
Permalink | Reply | 2 comments |
headers
Charles Sprickman | 1 Dec 2005 21:17
Picon
Favicon
Gravatar

Re: overloaded webserver: nfs wait issue?

On Thu, 1 Dec 2005, N.J. Thomas wrote:

> * N.J. Thomas <njt <at> ayvali.org> [2005-12-01 14:39:44 -0500]:
>> We have a website with moderately high traffic, load balanced among 3
>> webservers.
>>
>> During peak traffic times however (when the volume is higher than
>> normal), the load shoots up to over a 100, and the site crawls to its
>> knees.

The one thing that sticks out to me is that you go pretty deep into swap 
when the load shoots up.  Do you have room for more RAM?  What kind of 
hardware are we talking about here?

Charles

> I forgot to mention that the webservers are running FreeBSD 4.11 and
> Apache 1.3.x.
>
> Thomas
>
> -- 
> N.J. Thomas
> njt <at> ayvali.org
> Etiamsi occiderit me, in ipso sperabo
> _______________________________________________
> % NYC*BUG talk mailing list
> http://lists.nycbug.org/mailman/listinfo/talk
> %Be sure to check out our Jobs and NYCBUG-announce lists
> %We meet the first Wednesday of the month
(Continue reading)

Permalink | Reply | 1 comment |
headers
dlavigne6 | 1 Dec 2005 21:21
Picon
Favicon

python | snmp help


I finally have nav (that network monitoring application) up and running on a FreeBSD 6.0 system.
Everything works except for the editdb module which keeps failing with this error:

File "/usr/local/nav/lib/python/nav/Snmp.py", line 26, in ?
    import pysnmp.proto.cli.ucd

ImportError: No module named cli.ucd

>From the choices at freshports, I installed the packages that looked like they'd do the trick: py24-snmp
and ucd-snmp. I did get a whole bunch of psynmp subdirectories including proto, but it does not have any cli subdirectory...

Anyone have any clue how I can get cli.ucd on a BSD system?

Dru
Permalink | Reply | 0 comments |
headers
N.J. Thomas | 1 Dec 2005 21:27

Re: overloaded webserver: nfs wait issue?

* Charles Sprickman <spork <at> bway.net> [2005-12-01 15:17:38 -0500]:
> The one thing that sticks out to me is that you go pretty deep into swap 
> when the load shoots up.  Do you have room for more RAM?  What kind of 
> hardware are we talking about here?

We have:

    - one: P4, 3GHz, 1GB RAM
    - one: P3, 1.4GHz, 1GB RAM
    - two: P3, 900MHz, 1GB RAM

The two faster ones are used a bit more by the load balancer..

Thomas

--

-- 
N.J. Thomas
njt <at> ayvali.org
Etiamsi occiderit me, in ipso sperabo
Permalink | Reply | 0 comments |

Gmane