RE: Tomcat's scalability

2006-07-13 Thread Biernatowski Bartosz J
To follow up and close the Tomcat scalability thread I opened a while ago It turned out the bottleneck was on the database hardware end. Reducing disk utilization resolved our performance issues. BJ Biernatowski This e-mail, including attachments, may include confidential and/or proprietary i

Re: Tomcat's scalability

2006-06-21 Thread anjan bacchu
hi GB, From catalina.bat rem CATALINA_HOME May point at your Catalina "build" directory. rem rem CATALINA_BASE (Optional) Base directory for resolving dynamic portions rem of a Catalina installation. If not present, resolves to rem the same directory

Re: Tomcat's scalability

2006-06-21 Thread Andrew Miehs
Not only. The issue is really whether it make sense to use the operating systems process table as a queue or not. Up until linux new threading model in Linux kernel 2.6 this was definitely the case. The 2.6 threads are very efficient, so most programmers will not notice a real performance deg

Re: Tomcat's scalability

2006-06-21 Thread Alex Turner
This discussion focuses primarily on serving static files to a client, not processing dynamic web pages. Most people running tomcat are processing dynamic pages, like getting data from a database and compositing a page based on that data. An FTP site, or a static web site will typically be I/O b

Re: Tomcat's scalability

2006-06-21 Thread Andrew Miehs
Now that we are moving to the theoretical discussion, you will probably want to have a look at http://www.kegel.com/c10k.html Regards Andrew On 21/06/2006, at 4:56 PM, Mladen Adamovic wrote: I spoke recently with guy from Microsoft (project manager from server division). He said that hea

Re: Tomcat's scalability

2006-06-21 Thread Mladen Adamovic
Alex Turner wrote: Please also note that having a max threads of 750 is pretty much gaurtenteed to cause your system to grind to a halt under high load. (Most linux systems I've seen buckle somewhere around a load average of 75 or so, which means 75 threads waiting for CPU time). You mean 75

Re: Tomcat's scalability

2006-06-21 Thread Mladen Adamovic
Please see http://java.sun.com/developer/technicalArticles/Programming/linux/ Java on linux has been natively multithreaded since 1.3 Uops, I haven't known. Thank you all for your information (to Alex Turner, Leon Rosenberg, Darryl Milles). I was mistaken about this. -

Re: Tomcat's scalability

2006-06-21 Thread Darryl Miles
Mladen Adamovic wrote: Biernatowski, Is your HTTP application multi-threaded ? Irrelevant. Unimportant. Why is that ? What happens if his app is using this line in JSP ? <%@ page isThreadSafe="false" %> Google is your friend. Or to have i.e. extremely large Lucene database or some other

Re: Tomcat's scalability

2006-06-21 Thread Darryl Miles
Mladen Adamovic wrote: Max number of Java thread, IMHO. Java thread is not the same as operating system thread. In fact, JVM used to be single threaded on Linux and Windows and I'm not quite sure has it changed recently. So, you might have 800 Java threads but it is still one thread on operatin

Re: Tomcat's scalability

2006-06-20 Thread Alex Turner
Please see http://java.sun.com/developer/technicalArticles/Programming/linux/ Java on linux has been natively multithreaded since 1.3 Please also note that having a max threads of 750 is pretty much gaurtenteed to cause your system to grind to a halt under high load. (Most linux systems I've see

Re: Tomcat's scalability

2006-06-20 Thread Leon Rosenberg
On 6/21/06, Mladen Adamovic <[EMAIL PROTECTED]> wrote: Leon Rosenberg wrote: >> > Teoreticly, your servers should be faster if you configure 4 Tomcat >> > instances (4 JVMs) to do round robin. >> > You should be able to improve performances almost 4x. >> could you explain why?? >> I wanted to ask

Re: Tomcat's scalability

2006-06-20 Thread Mladen Adamovic
Leon Rosenberg wrote: > Teoreticly, your servers should be faster if you configure 4 Tomcat > instances (4 JVMs) to do round robin. > You should be able to improve performances almost 4x. could you explain why?? I wanted to ask the same question.. forget it. it was wrong. I think I wasn't wrong.

Re: Tomcat's scalability

2006-06-20 Thread Mladen Adamovic
Biernatowski Bartosz J wrote: I am about 90% sure the bottleneck is Tomcat or what's running on top of Tomcat. Application uses JDBC queries to MS SQL server Chips are Intel Xeon. My monitoring data: Memory utilization under 30%, CPU under 10%. Using hardcore performance tools and systematic app

Re: Tomcat's scalability

2006-06-20 Thread Mladen Adamovic
Darryl Miles wrote: LOL. Each HTTP request/response cycle is handed off to a worker thread, the available worker threads are dynamically increased to cope with the number of the simultaneous HTTP requests being processed in the moment. But isn't it Java threads. I'm speaking of operating sy

Re: Tomcat's scalability

2006-06-20 Thread Mladen Adamovic
Leon Rosenberg wrote: Isn't Tomcat and JVM still single threaded? Single thread = single processor usage I don't think it was ever singlethreaded. And if it were, what would the Connector setting in the server.xml mean? Max number of Java thread, IMHO. Java thread is not the same as operating

RE: Tomcat's scalability

2006-06-20 Thread Biernatowski Bartosz J
rol. > > > Basically, be warned: if you are running your connection to MS SQL Server > (and using the Microsoft JDBC drivers) through a firewall, this MAY be one > contributing factor of your poor performance. > > Just my $0.02. > > Gord > > > -----Original Messag

Re: Tomcat's scalability

2006-06-20 Thread Sérgio Costa
ne contributing factor of your poor performance. Just my $0.02. Gord -Original Message- From: Alex Turner [mailto:[EMAIL PROTECTED] Sent: Tuesday, June 20, 2006 12:42 AM To: Tomcat Users List Subject: Re: Tomcat's scalability On 6/19/06, Leon Rosenberg <[EMAIL PROTECTED]> wrote:

RE: Tomcat's scalability

2006-06-20 Thread Hyatt, Gordon
20, 2006 12:42 AM To: Tomcat Users List Subject: Re: Tomcat's scalability On 6/19/06, Leon Rosenberg <[EMAIL PROTECTED]> wrote: > > On 6/19/06, Biernatowski Bartosz J <[EMAIL PROTECTED]> wrote: > > I am about 90% sure the bottleneck is Tomcat or what's runnin

Re: Tomcat's scalability

2006-06-19 Thread frankpeng
install many many Tomcats. There is no scalability issue. Frank Peng. -Original Message- From: GB Developer <[EMAIL PROTECTED]> To: 'Tomcat Users List' Sent: Mon, 19 Jun 2006 16:40:17 -0500 Subject: RE: Tomcat's scalability How do you propose to add a 's

Re: Tomcat's scalability

2006-06-19 Thread Alex Turner
osenberg [mailto:[EMAIL PROTECTED] > Sent: Monday, June 19, 2006 10:49 AM > To: Tomcat Users List > Subject: Re: Tomcat's scalability > > are you sure that tomcat is your bottleneck? > Your 4 CPU machine (which cpu's btw?) should be able to handle more > than 1000 use

Re: Tomcat's scalability

2006-06-19 Thread Andrew Miehs
Ooops - forgot to add the rest Andrew Miehs wrote: Could be anything - the database could be the indexes in the database, could be deadlocks, could be a badly programmed application, could be high packet loss on the ethernet interfaces, could even be tomcat - As for the 90% guess - At l

Re: Tomcat's scalability

2006-06-19 Thread Leon Rosenberg
On 6/20/06, Andrew Miehs <[EMAIL PROTECTED]> wrote: Biernatowski Bartosz J wrote: > I am about 90% sure the bottleneck is Tomcat or what's running on top of > Tomcat. Application uses JDBC queries to MS SQL server > Chips are Intel Xeon. My monitoring data: If this is a REAL problem for

Re: Tomcat's scalability

2006-06-19 Thread Andrew Miehs
Biernatowski Bartosz J wrote: I am about 90% sure the bottleneck is Tomcat or what's running on top of Tomcat. Application uses JDBC queries to MS SQL server Chips are Intel Xeon. My monitoring data: Why are you 90% sure?! Your SQL server is running on a seperate machine? or the same machine?

RE: Tomcat's scalability

2006-06-19 Thread GB Developer
ty/clustering, so could be off. > -Original Message- > From: Biernatowski Bartosz J [mailto:[EMAIL PROTECTED] > Sent: Monday, June 19, 2006 4:55 PM > To: 'Tomcat Users List' > Subject: RE: Tomcat's scalability > > > My understanding of Tomcat&

Re: Tomcat's scalability

2006-06-19 Thread Leon Rosenberg
On 6/19/06, Biernatowski Bartosz J <[EMAIL PROTECTED]> wrote: On 6/19/06, Mladen Adamovic <[EMAIL PROTECTED]> wrote: > Do top on the servers to be sure is problem in Tomcat or not. > > Teoreticly, your servers should be faster if you configure 4 Tomcat > instances (4 JVMs) to do round robin. > Yo

RE: Tomcat's scalability

2006-06-19 Thread Biernatowski Bartosz J
On 6/19/06, Mladen Adamovic <[EMAIL PROTECTED]> wrote: > Do top on the servers to be sure is problem in Tomcat or not. > > Teoreticly, your servers should be faster if you configure 4 Tomcat > instances (4 JVMs) to do round robin. > You should be able to improve performances almost 4x. > could you

Re: Tomcat's scalability

2006-06-19 Thread Leon Rosenberg
ing else is just kindergarten :-) BJ Biernatowski Application Developer, e-Business Leon -Original Message- From: Leon Rosenberg [mailto:[EMAIL PROTECTED] Sent: Monday, June 19, 2006 10:49 AM To: Tomcat Users List Subject: Re: Tomcat's scalability are you sure that tomcat is your b

RE: Tomcat's scalability

2006-06-19 Thread Biernatowski Bartosz J
2:40 PM To: 'Tomcat Users List' Subject: RE: Tomcat's scalability How do you propose to add a 'separate instance of Tomcat' without 'adding a separate JVM'? Or do you/others mean by 'instance of tomcat' = 'a separate physical server with single in

RE: Tomcat's scalability

2006-06-19 Thread GB Developer
How do you propose to add a 'separate instance of Tomcat' without 'adding a separate JVM'? Or do you/others mean by 'instance of tomcat' = 'a separate physical server with single instance of JVM/Tomcat' ? > > So far it sounds that the approach of adding separate > instance of Tomcat and using

Re: Tomcat's scalability

2006-06-19 Thread Alex Turner
06 10:49 AM To: Tomcat Users List Subject: Re: Tomcat's scalability are you sure that tomcat is your bottleneck? Your 4 CPU machine (which cpu's btw?) should be able to handle more than 1000 users (unless you are speaking about suns cpu) without problems. Maybe you should provide more info

Re: Tomcat's scalability

2006-06-19 Thread Alex Turner
10:49 AM To: Tomcat Users List Subject: Re: Tomcat's scalability are you sure that tomcat is your bottleneck? Your 4 CPU machine (which cpu's btw?) should be able to handle more than 1000 users (unless you are speaking about suns cpu) without problems. Maybe you should provide more info abo

RE: Tomcat's scalability

2006-06-19 Thread Biernatowski Bartosz J
pplication Developer, e-Business -Original Message- From: Leon Rosenberg [mailto:[EMAIL PROTECTED] Sent: Monday, June 19, 2006 10:49 AM To: Tomcat Users List Subject: Re: Tomcat's scalability are you sure that tomcat is your bottleneck? Your 4 CPU machine (which cpu's btw?) shou

Re: Tomcat's scalability

2006-06-19 Thread Darryl Miles
Leon Rosenberg wrote: On 6/19/06, Mladen Adamovic <[EMAIL PROTECTED]> wrote: Leon Rosenberg wrote: >> Teoreticly, your servers should be faster if you configure 4 Tomcat >> instances (4 JVMs) to do round robin. >> You should be able to improve performances almost 4x. >> > could you explain why??

Re: Tomcat's scalability

2006-06-19 Thread Leon Rosenberg
On 6/19/06, Mladen Adamovic <[EMAIL PROTECTED]> wrote: Leon Rosenberg wrote: >> Teoreticly, your servers should be faster if you configure 4 Tomcat >> instances (4 JVMs) to do round robin. >> You should be able to improve performances almost 4x. >> > could you explain why?? Isn't Tomcat and JVM s

Re: Tomcat's scalability

2006-06-19 Thread Mladen Adamovic
Leon Rosenberg wrote: Teoreticly, your servers should be faster if you configure 4 Tomcat instances (4 JVMs) to do round robin. You should be able to improve performances almost 4x. could you explain why?? Isn't Tomcat and JVM still single threaded? Single thread = single processor usage ---

Re: Tomcat's scalability

2006-06-19 Thread Leon Rosenberg
On 6/19/06, Mladen Adamovic <[EMAIL PROTECTED]> wrote: Do top on the servers to be sure is problem in Tomcat or not. Teoreticly, your servers should be faster if you configure 4 Tomcat instances (4 JVMs) to do round robin. You should be able to improve performances almost 4x. could you explai

Re: Tomcat's scalability

2006-06-19 Thread Mladen Adamovic
Do top on the servers to be sure is problem in Tomcat or not. Teoreticly, your servers should be faster if you configure 4 Tomcat instances (4 JVMs) to do round robin. You should be able to improve performances almost 4x. Biernatowski Bartosz J wrote: Hello, I was hoping somebody on the list

Re: Tomcat's scalability

2006-06-19 Thread Leon Rosenberg
Almost forgot, as for your question about multiple jvms with multiple tomcat instances: we tried to scale tomcat instances on the same machine and it made no difference. leon On 6/19/06, Biernatowski Bartosz J <[EMAIL PROTECTED]> wrote: Hello, I was hoping somebody on the list might point me i

Re: Tomcat's scalability

2006-06-19 Thread Leon Rosenberg
are you sure that tomcat is your bottleneck? Your 4 CPU machine (which cpu's btw?) should be able to handle more than 1000 users (unless you are speaking about suns cpu) without problems. Maybe you should provide more info about your application. Do you have any monitoring data? Leon On 6/19/06,

Tomcat's scalability

2006-06-19 Thread Biernatowski Bartosz J
Hello, I was hoping somebody on the list might point me in the right direction... I am trying to scale up Tomcat based web application currently supporting ~100 users to 350 users. It seems that I have enough hardware: 2 load balanced servers x 4 CPUs each with 4 GB of RAM which is underutilized