Hi!
Thanks for your reply!
If Java and Tomcat is multi-threaded applications, I still don't understand why I see 
Java on only one CPU when I use the top function in Linux. Here is a snapshot from a 
medium stress-test load:

 15:02:23  up 1 day, 22:23,  1 user,  load average: 1,79, 1,56, 0,74
82 processes: 76 sleeping, 5 running, 1 zombie, 0 stopped
CPU states:  cpu    user    nice  system    irq  softirq  iowait    idle
           total   50,0%    0,0%    7,4%   0,0%     4,6%    0,5%   37,2%
           cpu00   31,9%    0,0%    8,9%   0,0%     5,9%    1,1%   51,9%
           cpu01   68,1%    0,0%    5,9%   0,0%     3,3%    0,0%   22,5%
Mem:  1286812k av,  887436k used,  399376k free,       0k shrd,  162608k buff
                    462864k actv,  193060k in_d,     256k in_c
Swap: 3072216k av,       0k used, 3072216k free                  392952k cached

  PID USER     PRI  NI  SIZE  RSS SHARE STAT %CPU %MEM   TIME CPU COMMAND
  745 root      25   0 98,4M  98M  7400 S    42,7  7,8 158:48   0 java
 6048 mysql     15   0  101M 101M  2576 S     0,9  8,0   1:12   1 mysqld
 6049 mysql     25   0  101M 101M  2576 R     0,6  8,0   1:49   1 mysqld
13138 root      15   0  1292 1292   904 R     0,3  0,1   0:03   0 top
 6400 mysql     25   0  101M 101M  2576 S     0,1  8,0   0:00   0 mysqld
 6410 mysql     25   0  101M 101M  2576 S     0,1  8,0   0:00   0 mysqld
 6528 mysql     25   0  101M 101M  2576 S     0,1  8,0   0:00   0 mysqld
    1 root      15   0   500  500   444 S     0,0  0,0   0:08   0 init
    2 root      RT   0     0    0     0 SW    0,0  0,0   0:00   0 migration/0
    3 root      RT   0     0    0     0 SW    0,0  0,0   0:00   1 migration/1
    4 root      15   0     0    0     0 SW    0,0  0,0   0:00   0 keventd
    5 root      34  19     0    0     0 SWN   0,0  0,0   0:00   0 ksoftirqd/0
    6 root      34  19     0    0     0 SWN   0,0  0,0   0:00   1 ksoftirqd/1
    9 root      25   0     0    0     0 SW    0,0  0,0   0:00   0 bdflush
    7 root      15   0     0    0     0 SW    0,0  0,0   0:02   0 kswapd
    8 root      15   0     0    0     0 SW    0,0  0,0   0:03   0 kscand
   10 root      15   0     0    0     0 SW    0,0  0,0   0:04   0 kupdated
   11 root      25   0     0    0     0 SW    0,0  0,0   0:00   0 mdrecoveryd
   20 root      15   0     0    0     0 SW    0,0  0,0   0:01   0 kjournald
   75 root      25   0     0    0     0 SW    0,0  0,0   0:00   1 khubd
  145 root      15   0     0    0     0 SW    0,0  0,0   0:05   1 kjournald
  146 root      15   0     0    0     0 SW    0,0  0,0   0:00   0 kjournald
  147 root      15   0     0    0     0 SW    0,0  0,0   0:11   0 kjournald
  148 root      15   0     0    0     0 SW    0,0  0,0   0:02   1 kjournald
  492 root      15   0   568  568   492 S     0,0  0,0   0:00   1 syslogd
  496 root      25   0   448  448   392 S     0,0  0,0   0:00   0 klogd
  506 root      15   0   444  444   384 S     0,0  0,0   0:23   1 irqbalance
  523 rpc       25   0   560  560   488 S     0,0  0,0   0:00   0 portmap
  542 rpcuser   25   0   716  716   636 S     0,0  0,0   0:00   0 rpc.statd
  622 root      15   0  1460 1460  1228 S     0,0  0,1   0:02   0 sshd
  636 root      25   0   792  792   684 S     0,0  0,0   0:00   1 xinetd
  646 root      15   0   488  488   408 S     0,0  0,0   0:00   0 vsftpd
  667 root      15   0  2584 2584  1892 S     0,0  0,2   0:07   0 sendmail
  676 smmsp     15   0  2260 2252  1708 S     0,0  0,1   0:00   1 sendmail
  754 root      15   0   464  464   408 S     0,0  0,0   0:00   0 gpm
  763 root      15   0   612  612   544 S     0,0  0,0   0:00   0 crond
  806 daemon    15   0   556  556   492 S     0,0  0,0   0:00   0 atd
  842 root      15   0  4972 4972  1900 S     0,0  0,3   0:07   1 miniserv.pl
  846 root      15   0   416  416   368 S     0,0  0,0   0:00   1 mingetty
  847 root      15   0   420  420   368 S     0,0  0,0   0:00   0 mingetty
  848 root      15   0   420  420   368 S     0,0  0,0   0:00   1 mingetty
  849 root      21   0   420  420   368 S     0,0  0,0   0:00   0 mingetty
  850 root      17   0   420  420   368 S     0,0  0,0   0:00   1 mingetty
  851 root      17   0   416  416   368 S     0,0  0,0   0:00   1 mingetty
 6018 root      25   0  1128 1128   984 S     0,0  0,0   0:00   1 mysqld_safe
 6050 mysql     20   0  101M 101M  2576 S     0,0  8,0   0:00   0 mysqld
 6051 mysql     15   0  101M 101M  2576 S     0,0  8,0   0:00   0 mysqld
 6052 mysql     15   0  101M 101M  2576 S     0,0  8,0   0:00   0 mysqld
 6053 mysql     15   0  101M 101M  2576 S     0,0  8,0   0:10   0 mysqld
 6054 mysql     15   0  101M 101M  2576 S     0,0  8,0   0:00   1 mysqld
 6055 mysql     15   0  101M 101M  2576 S     0,0  8,0   1:17   1 mysqld
 6056 mysql     15   0  101M 101M  2576 S     0,0  8,0   0:29   0 mysqld
 6057 mysql     15   0  101M 101M  2576 S     0,0  8,0   0:09   1 mysqld
13091 root      15   0  2084 2084  1716 S     0,0  0,1   0:00   1 sshd
13093 root      23   0  1336 1336  1128 S     0,0  0,1   0:00   0 bash
 6369 mysql     25   0  101M 101M  2576 S     0,0  8,0   0:00   1 mysqld
 6385 mysql     25   0  101M 101M  2576 S     0,0  8,0   0:00   1 mysqld

As you see, there's only one Java service on CPU 0, and MySQL is running on both CPUs.
Any explanation for this?
The web application running is a web publishing system based on web contents stored in 
a MySQL database. So there's lots of SQL queries, but fron the info above, java seems 
to be the bottleneck in the system.

Any help to increase the performance is appreciated.

Thanks
Jens Ove


----- Original Message ----- 
From: "Peter Guyatt" <[EMAIL PROTECTED]>
To: "Tomcat Users List" <[EMAIL PROTECTED]>; "Jens Ove Lillegraven" <[EMAIL PROTECTED]>
Sent: Wednesday, November 12, 2003 4:31 PM
Subject: RE: How to tune Tomcat and Java?


> Hi There
> 
> sorry sent the last one before it was ready ;-)
> 
> Java and Tomcat are both multi-threaded applications. So they are both
> using more than one CPU (if available).
> 
> If it is using 50% of one processor then it is using 100% CPU on that
> processor, since its 50% of the actual processing capacity.
> 
> Check for an infinate loop without a sleep in your code, especially if you
> are attempting to connect a client socket to a server since linux does not
> wait before throwing an exception.
> 
> Thanks
> 
> Pete
> 
> -----Original Message-----
> From: Jens Ove Lillegraven [mailto:[EMAIL PROTECTED]
> Sent: 12 November 2003 15:17
> To: [EMAIL PROTECTED]
> Subject: How to tune Tomcat and Java?
> 
> 
> Hi,
> 
> We have a problem in that the Tomcat (java) process is the bottleneck and is
> using only one CPU.
> There will be high load on the server when it is put online, so we are
> stress testing it offline now. Using the top command in Linux during the
> stress test, it seems that java is the bottleneck of the system. But it uses
> not 100%, but about 45-50% of one CPU. We get no http error, the page
> performance (pages pr. second) is just getting worse as we put on more
> stress load. The max memory consumption is set to 300M, but I've never seen
> java using more than 111M. I suppose the throughput would have increased if
> both CPUs could be used, but is Java (Tomcat) a single processor
> application? The network throughput has a maximum average of 500kbps
> (kilo-bits) during the stress test, and should manage much more.
> Anyone who know how to make java and Tomcat performing better?
> 
> Details of hardware and innstalled applications:
> Compaq ProLiant 6000, 2 x PII Xeon 400MHz, 1,3GB RAM
> Red Hat Linux Enterprise 3 ES
> J2EE: j2sdkee-1_3_1-linux
> J2SE: j2sdk-1_4_2_02-linux-i586
> Tomcat: 4.1.24
> MySQL: 4.0.16
> 
> Tuning Tomcat memory (done):
> in /usr/bin/dtomcat4:
>   CATALINA_OPTS="-Xmx300M -Xms100M $CATALINA_OPTS $JPDA_OPTS"
> 
> MySQL has been tuned, and is running very well.
> 
> Thanks
> Jens Ove
> 
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
> 
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to