Messaging performance problem on RAID disk with many IMAP users

The performance problem can be solved by tunning the disc RAID system.

Similar Messages

I have several disks with many comedy routines;I want to select a few from each disk and combine them to one disk for my personal use.How?

I have several disks with many comedy routines;I want to select a few from each disk to make one single disk for my personal use.How?

Copy the routines into a new folder on your hard drive. Then burn the folder to a new blank disc using Disk Utility.

Performance Problem ADF-UIX Appl. with Bea 8.1

Hello JDev & J2EE Folks,
I'm running into a performance problem with an ADF-UIX Application (JDev Version 10.1.2.0.0) under Bea 8.1. Application Server. It seems that the the same application is running at least factor 5 slower than on OC4J. I didn't experience such a difference with the older JDev9i Applications, seems to be an issue with 10g. Not done much investigation so far. Are there any ideas or suggestions out there? Anyone who had a similar problem?

Hi ,
Did you get this resolved? Can you share? Thanks!

Performance problem at bulk insert with spatial index

Hi,
I have a table with SDO_GEOMETRY.
Insert without spatial index is very fast, but with active spatial index it's very slow.
So for the first big import of data, I can drop the index, import the data and again create the index. Thats 10 times faster!
But for an already very big table that is no option.
The 10g1-Users Guide (1) says at 4.1.3 that the spatial index should be set to 'deferred', the data should be inserted and than the index should be synchronized again. That sounds very good, but I can't find this at the 11g1-Users Guide.
I tried it (11g1), but the performance is even worse than with active index!
What could be my mistake? Any hints?
Thank you,
Bjoern Weitzig
create table sggeoptcollection (pt SDO_GEOMETRY);
CREATE INDEX myIdx ON sggeoptcollection (pt) INDEXTYPE IS MDSYS.SPATIAL_INDEX PARAMETERS('sdo_indx_dims=2, layer_gtype=point, sdo_rtr_pctfree=50');
ALTER INDEX myIdx PARAMETERS ('index_status=deferred');
Big import with batch'ed PreparedStatements
ALTER INDEX myIdx PARAMETERS ('index_status=synchronize sdo_batch_size=500');
1) http://download.oracle.com/docs/html/B10826_01/sdo_index_query.htm#g1010227

Hi,
I have a table with SDO_GEOMETRY.
Insert without spatial index is very fast, but with active spatial index it's very slow.
So for the first big import of data, I can drop the index, import the data and again create the index. Thats 10 times faster!
But for an already very big table that is no option.
The 10g1-Users Guide (1) says at 4.1.3 that the spatial index should be set to 'deferred', the data should be inserted and than the index should be synchronized again. That sounds very good, but I can't find this at the 11g1-Users Guide.
I tried it (11g1), but the performance is even worse than with active index!
What could be my mistake? Any hints?
Thank you,
Bjoern Weitzig
create table sggeoptcollection (pt SDO_GEOMETRY);
CREATE INDEX myIdx ON sggeoptcollection (pt) INDEXTYPE IS MDSYS.SPATIAL_INDEX PARAMETERS('sdo_indx_dims=2, layer_gtype=point, sdo_rtr_pctfree=50');
ALTER INDEX myIdx PARAMETERS ('index_status=deferred');
Big import with batch'ed PreparedStatements
ALTER INDEX myIdx PARAMETERS ('index_status=synchronize sdo_batch_size=500');
1) http://download.oracle.com/docs/html/B10826_01/sdo_index_query.htm#g1010227

Performance problem on a table with zero rows

I queried the v$sqlarea table to find which SQL statement was doing the most disk reads and it turned out to be a query that read 278 Gigabytes from a table which usually has zero rows. This query runs every minute and reads any rows that are in the table. It does not do any joins. The process then processes the rows and deletes them from the table. Over the course of a day, a few thousand rows may be processed this way with a row length of about 80 bytes. This amounts to a few kilobytes, not 278 Gig over a couple of days. Buffer gets were even higher at 295 Gig. Note that only the query that reads the table is doing these disk reads, not the rest of the process.
There are no indexes on the table, so a full table scan is done. The query reads all the rows, but usually there are zero. I checked the size of the table in dba_segments, and it was 80 Meg. At one point months ago, during a load, the table had 80 Meg of data in it, but this was deleted after being processed. The size of the table was never reduced.
I can only assume that Oracle is doing a full table scan on all 80 Meg of this table every minute. Even when there are zero rows. Dividing the buffer gets in bytes by the number of executions yields 72 Meg which is close to the table size. Oracle is reading the entire table size from disk even when the table has zero rows.
The solution was to truncate the table. This helped immediately and reduced the disk reads to zero most of the time. The buffer gets were also reduced to 3 per execution when the table was empty. The automatic segment manager reduced the size of the table to 64k overnight.
Our buffer cache hit ratio was a dismal 72%. It should go up now that this problem has been resolved.
Table statistics are gathered every week. We are running Oracle 9.2 on Solaris.
Note that this problem is already resolved. I post it because it is an interesting case.
Kevin Tyson, OCP
DaimlerChrysler Tech Center, Auburn Hills, MI

Kevin,
The solution was to truncate the tableThis is not a scoop... isn't it ?
Table statistics are gathered every weekIs there any reason for that ?
If stats ran when no rows, perf can be very bad after loading data, and if stats ran when thousand rows, perf can be very bad after deleting. Perhaps can you find a more restrictive stat running ?
Nicolas.
Message was edited by:
N. Gasparotto

Problem Rebuilding RAID set with 10.5?

I have 1 RAID set (mirrored with autorebuild) composed of 2 (identical) HD. I got a message in 10.5 Disk Utility that one disk had failed and that the set was degraded.
I replaced the drive and I am getting an error message during the rebuild "Unrecognized filesystem".
I can still see and access my RAID set (1 slice still ok) so minimal damage here but I don't understand why I am getting this error message when trying to rebuild.
Any idea?
Thanks
setup:
-mirrored raid set
- autorebuild selected
- 2 slices (160Gb, GUI, OS journaled ext)

Hi VK2CPJ,
Today ( I slept on this idea was exhausted from previous day) tried on two blank eide drives in a raid box, no raid built just two independent drives, one set for master the other slave. A couple of observations:
I did not raid because I want one small partition to contain a Leopard boot install from the DVD. The message I received from disk utility was that if I restore that DVD my raid may be unusable(if I did raid). So I did not raid the drives but let them stand as master and slave. I wanted to do a APM on the small "boot" partition to see if my intel and PPC Mac would in fact boot from that restored Leopard DVD boot partition. I used start up pane on both PPC and intel and they would not boot up from that Leopard partition even though the start up pane showed the disk as a startup device. Using the option key booted both Macs. Now I'm not sure if doing this was to procedure because the Apple docs says to use a PPC Mac to do this process. Which I have a difficult time to understand. I'm assuming that the doc says that because APM would be default and assisting the user not to make any option mistake. So currently I'm trying this same procedure using a PPC Mac to test out whether start up pane will boot both Macs. I don't think it will but do think using option within the start up manager will work. If so then does this mean imply that the start up pane has a bug or that user does not require a PPC Mac to do the setup process. Let you know shortly.

Performance problems using WLS6.1 with the thin driver for Ora9i from OTN

Hello everyone,
Has anyone been experimenting with the thin driver for Oracle 9i under WLS6.1. I got it running, but
I get extremely bad performance when having multiple concurrent accesses. It works more or less in
single user mode. Also it seems not to support XA transactions?
When could we expect the driver from BEA to be out?
Thanks in advance!
Samuel Kounev
TU-Darmstadt, Germany

I had done this from the very beginning, but I was still getting very bad performance under
concurrent accesses. I even followed Sree's instructions and removed the old thin driver from
weblogic.jar, but this didn't help. I am now trying to get the OCI driver for 9i running and hope
this will solve the problem.
Samuel
Tamilselvan Ramasamy wrote:
You have to use Oracle's Oracle 9i thin driver .
put this driver infront of the classpath i.e before weblogic_sp.jar and
weblogic.jar
/selvan
"Sree Bodapati" <[email protected]> wrote in message
news:[email protected]..
The BEA driver is available in WLS6.1SP2 which is GA.
hth
sree
"Samuel Kounev" <[email protected]> wrote in message
news:[email protected]..
Hello everyone,
Has anyone been experimenting with the thin driver for Oracle 9i under
WLS6.1. I got it running, but
I get extremely bad performance when having multiple concurrent accesses.
It
works more or less in
single user mode. Also it seems not to support XA transactions?
When could we expect the driver from BEA to be out?
Thanks in advance!
Samuel Kounev
TU-Darmstadt, Germany

How to mount RAID disk with target mode

Hi,
On my Mac Pro I have 2x2TB HHD which I use as RAID 1.
For maintenance I need now to access the HHD content from my macbook via FIreWIre target mode. On the macbook however I see the two HHD unmounted as two independent units, not as RAID 1.
How can I safely mount my RAID 1 disk?

I don't think you'll need another computer or eight hours of downtime. You didn't mention the size of your internal drive but I'd assume the 300gb external will hold it all. I'll also assume the external is a firewire drive, if it's not none of this will work.
This is what I would do:
1. Download Carbon Copy Cloner or SuperDuper to clone your remaining raid slice to the external 300gb drive, and make sure to check the "make bootable" box.
2. In System Preferences set your startup disk to the external.
3. Restart from the external.
4. Run disk Utility to reformat your internal drive. At this point you're going to need to decide wether or not to set it up as a raid again. I don't think I would.
5. Use Carbon Copy Cloner or SuperDuper to clone the external back onto your internal drive.
6. System Prefs again to reset the startup drive to your internal.
7. Restart and you should be good to go, and you've got your first backup done on your external.
8. Devise and impliment a backup strategy. Do you need daily/weekly/monthly backups? You'll have to decide how much you can afford to lose if your main drive goes belly up. There are several programs that are easy to set up a backup schedule on, the above mentioned CCC and SuperDuper and a little app called Deja Vu which runs as a system preference pane, is very easy to set up and if I remember is free.
Now, I say all of this not knowing what you do with your external drive to begin with. I am guessing that the raid was set up as a mirror to be a backup if the drive had problems....which is exactly where you are. Personally I would just format it as one drive and backup to the external. If the internal does actually fail you'll still have all your stuff and can run from the external until you replace the internal.
Make sense? I know I rambled a bit but it's really easy.
Dual 2.0, 4 GB, 2x160 GB , SuperDrive, 9800 SE, 20 Aluminum - 15 1.5 PB, Mac OS X (10.4.5)

Problems after repairing disk with Mac OS X install Discs

Having been experiencing slowness in my iMac, I tried to verify disk and found that my HD Volume needed to be repaired. However, the "Repair Disk" button was gray. I later learned from online mac discussions that I should reboot my iMac with the Mac OSx install Disc. I followed the steps and then got the answer that there was nothing to be repaired.
The next thing I knew was that I could not get out of there. By choosing "Restart", I was bought back to choosing the language and installing the disc. Not wanting to do that I went around and around and finally had no choice but installing the Disc 1 and then 2. At one point, I was asked if I wanted to move previous files to the new location (don't remember the exact words), I said no. I was astonished to find out very soon that my computer was put back at its original state, meaning all the applications were empty.
I managed to get my safari and email (Microsoft Entourage) back. I transfered all photos from the "Previous System" folder to the present iPhoto. However, they are only sorted by year. all the categories and sub-categories that I so painstakingly created were gone, so were the keywords.
Among other problems, my iTune wouldn't open unless I have Mac OS X 10.4.9 (mine is 10.4.4). I am quite desperate at this moment. I know the previous settings are somewhere but have no idea how to bring them back.
I hope someone can help me get out of this mess. Thank in advance.
P.S. my iMac is much faster now. That is the only good news.
LN

There is an option to +select the drive to restart from+ (don't remember the exact wording) in the Installer's menu. If you previously set the startup disk to be the disc in the optical drive, it will continue to restart from the Mac OS X installation disk until you change it. Alternately, you could press and hold the Option key at startup to bring up Startup Manager. This will allow you to select the disk to use for startup. For future reference...
If you did a complete reinstallation of the OS, you need to select Software Update from the Apple menu. This will get to you to the latest version of Tiger, which is 10.4.11.

Problems using new version with many sites including facebook...how do I uninstall it NOW

Can't access a multi-list site for my business as a real estate agent...totally screwed up my chats on facebook...I want to uninstall this new version

-> [[Uninstalling Firefox]] - Uninstall your current Firefox
* REMOVE USER DATA AND SETTINGS
* Restart your system
-> Download latest Firefox version from one of these links using some other Web Browser:
* http://www.mozilla.com/en-US/firefox/new/
* http://www.mozilla.com/en-US/firefox/central/
* http://www.mozilla.com/en-US/firefox/fx/
* http://www.mozilla.com/en-US/firefox/all.html
-> [[Installing Firefox]]
Check and tell if its working.

CQ Sites with many registered users

Hi
With CQ 5.2.1 I tried to setup a mass-user site registering users as standard CQ users --> more than 50.000 users.
I soon discovered that user creation started to slow a lot (40-50 seconds to create a new user) and decided then to manage users in a different way.
Analyzing the node structure for users:
/home/users/<first_letter_username>/**** 1 folder for each username
It's clear that with thousands of nodes, childs of <first_letter_username> folders will grow a lot. If there is one thing I have understood with JCR is: 'please don't put too many childs in the same folder'. A new node creation tends to become pretty slow, near blocking.
On that solution (5.2.1) I moved to a database approach to manage my user data and authentication. Now I'm facing similar scenarios on 5.4 instance and would like to use embedded CQ user management
My question... the above mentioned user structure in the repository is still there in CQ 5.4. Am I going toward performance issues when creating thousands of users?
What about newsletter registrations? if I want to use newsletters in MCM I'm going to insert many users in these folders. Are there any performance issues I might consider? Or best practices to follow?
thanks

Hi,
Planning to work with a site with a lot of users as well. Did you use a database or LDAP on CQ 5.4 to manage a large number of users?
Thanks,

Performance Problems with intrinsic array multiplikation in Fortran with Su

Hello,
I found a serious performance problem in my application with intrinsic array multiplikation. Therefore I wrote a small test program to check if this a general problem or only exists in my code.
The test program (as seen below) was compiled with Sunstudio12 und Solaris 5.10 on an 64 bit Amd Opteron Machine. First with high optimization (f95 -fast -g), and the second time with out optimization (f95 -O0 -g). In both cases the intrinsic array mupltiplication had a lot of tlb and L2 cache misses (I made some test with the performance analyzer) which caused a huge increase in computation time compared to the explicit loop. In the ideal case the compiler should create the nested loops from the intrinsic statement and both functions should use exactly the same computing time.
I also tried compiling with Studio11, the problem also occurs there. Maybe its a compiler bug but I'm not sure.
Greetings
Michael Kroeger
program test
!Test zur Ausfuehrung einer einfachen array Multiplikation
implicit none
real,dimension(:,:,:),pointer::check1,check2
integer i,j,k,ni,nj,nk,status
ni=1000
nj=1000
nk=100
allocate(check1(0:ni,0:nj,0:nk),STAT=status)
write(*,*)status
allocate(check2(0:ni,0:nj,0:nk),STAT=status)
write(*,*)status
check1(i,j,k)=25
check2(i,j,k)=25
call intrinsic_f(check1)
call withloop(check2,i,j,k)
deallocate(check1,check2)
contains
subroutine intrinsic_f(check1)
real, dimension(:,:,:),pointer::check1
check1=check1*5
end subroutine intrinsic_f
subroutine withloop(check2,ni,nj,nk)
real, dimension(:,:,:),pointer::check2
integer i,j,k,nk,nj,ni
do k=0,nk
do j=0,nj
do i=0,ni
check2(i,j,k)=check2(i,j,k)*5
enddo
enddo
enddo
end subroutine withloop
end program

I will try the large pages, but what me puzzles is, that the intrinsic function has cache misses, but the loop function has not. All documentation says, that the compiler would expand the intrinsic function into 3 nested loops, so that both functions should be essential the same (and the compiler says it has created 3 loops from the intrinsic). Since they are not the same, this is a severe problem.
I have a fairly large code for high performance simulations, and this intrinsic functions are a significant bottle neck. If it is not a compiler bug, I would have to rewrite all intrinsic functions into loops.

Performance problem in RFC to JDBC interface

Hello everybody!
i'm working whit SAP PI 7.1
We defined some interfaces RFC - PI - JDBC (SQL server) but we have some performance problem.
If we have many row to write on the table then interface finish in timeout :
Synchronous timeout exceeded.
Returning to application. Exception: com.sap.engine.interfaces.messaging.api.exception.MessageExpiredException: Message 1d1f00b0-fecf-11de-8738-0015600446f0(OUTBOUND) expired.
I read the PI tuning document and i tried to apply configuration whit Advanced Adapter Engine but whitout result.
Now we want change the timeout in visual admin and maybe we solve the error but i'm asking myself....:
It's normal that for write 1500 row in a table we need more than 4 minuts????
It's possible accelerate this process??? After go live we will write messages whit more than 50.000 row.
somebody may help me?
PS: please no link to tuning guide or to notes (to increase the timeout parameter).

This could be because your Database system (JDBC server) is taking more time to insert. The problem is not on PI side but on the receiving system side. Try inserting the same number od rows on the database server itself and check for the time taken for execution. Adding indexes on your database table solves the issue lot of times.
Here PI is not the culprit but definitely the receiver system.
VJ

Performance problem (massive imap getUnreadMessageCount requests)

Hello,
I'm working on a server that's supposed to access the unread message count of many IMAP users (up to 40/second).
I'm facing performance issues : i can't get more than ~20/second.
I tried using multiple stores (and sessions) but it won't go faster. The performance bottleneck seems to be in the "connect" part.
I can provide the server configuration files if needed.
Maybe there's performance tuning I'm not aware of that I could use, can you help?
A very basic program that gets called way too often :
import java.util.Date;
import java.util.Properties;
import javax.mail.Folder;
import javax.mail.Session;
import com.sun.mail.imap.IMAPStore;
public class MessagesDao {
* singleton
private static MessagesDao instance = null;
* singleton
* @return
public static synchronized MessagesDao getInstance() {
if (instance == null) {
instance = new MessagesDao();
return instance;
private Session sessionImap;
* MessagesDao
private MessagesDao() {
final String portServeurImap = "143";
final String adresseServeurImap = "my_server";
final String imapConnectionPoolSize = "10";
final String imapConnectionPoolTimeout = "10";
final String imapTimeout = "2000";
final String imapConnectionTimeout = "2000";
final Properties propSessionImap = new Properties();
propSessionImap.put("mail.store.protocol", "imap");
propSessionImap.put("mail.imap.connectionpoolsize", imapConnectionPoolSize);
propSessionImap.put("mail.imap.connectionpooltimeout", imapConnectionPoolTimeout);
propSessionImap.put("mail.imap.timeout", imapTimeout);
propSessionImap.put("mail.imap.connectiontimeout", imapConnectionTimeout);
propSessionImap.put("mail.imap.port", portServeurImap);
propSessionImap.put("mail.imap.host", adresseServeurImap);
propSessionImap.put("mail.imap.auth", "true");
propSessionImap.put("mail.imap.separatestoreconnection", "true");
propSessionImap.put("mail.imap.appendbuffersize", -1);
propSessionImap.put("mail.debug", "false");
this.sessionImap = Session.getInstance(propSessionImap);
* user : someName@myDomain
public int getUnreadMessageCount(String user){
IMAPStore store = null;
try {
long timeStart = new Date().getTime();
store = (IMAPStore) this.sessionImap.getStore();
long timeGetStore = new Date().getTime();
store.connect(user, "fakepass");
long timeConnect = new Date().getTime();
Folder inbox = store.getFolder("inbox");
int count = inbox.getUnreadMessageCount();
System.out.println("-----> unread " + user + " : " + count + "( getStore : " + (timeGetStore - timeStart) + ", connect : " + (timeConnect - timeGetStore)+")");
return count;
} catch (Exception e) {
e.printStackTrace();
return -1;
} finally {
try {
store.close();
} catch (Exception e) {
Edited by: user13568923 on 29 déc. 2010 09:46

A single user never opens more than 2 folders at a time. I guess I don't really need the 10 folder pool...
Here's what the traffic looks like :
"5388","16.536513","cli","serv","TCP","46434 > imap [SYN] Seq=0 Win=5840 Len=0 MSS=1460 TSV=58648306 TSER=0 WS=6"
"5389","16.537545","serv","cli","TCP","imap > 46434 [SYN, ACK] Seq=0 Ack=1 Win=5792 Len=0 MSS=1460 TSV=3608644264 TSER=58648306 WS=7"
"5390","16.537567","cli","serv","TCP","46434 > imap [ACK] Seq=1 Ack=1 Win=5888 Len=0 TSV=58648306 TSER=3608644264"
"5391","16.539219","serv","cli","IMAP","Response: * OK Dovecot ready."
"5392","16.539251","cli","serv","TCP","46434 > imap [ACK] Seq=1 Ack=22 Win=5888 Len=0 TSV=58648307 TSER=3608644266"
"5394","16.540835","cli","serv","IMAP","Request: A0 CAPABILITY"
"5397","16.542140","serv","cli","TCP","imap > 46434 [ACK] Seq=22 Ack=16 Win=5888 Len=0 TSV=3608644269 TSER=58648307"
"5398","16.542399","serv","cli","IMAP","Response: * CAPABILITY IMAP4rev1 SASL-IR SORT THREAD=REFERENCES MULTIAPPEND UNSELECT LITERAL+ IDLE CHILDREN NAMESPACE LOGIN-REFERRALS QUOTA STARTTLS AUTH=PLAIN"
"5399","16.542733","cli","serv","IMAP","Request: A1 AUTHENTICATE PLAIN"
"5415","16.584176","serv","cli","TCP","imap > 46434 [ACK] Seq=202 Ack=39 Win=5888 Len=0 TSV=3608644310 TSER=58648307"
"5421","16.594994","serv","cli","IMAP","Response: + "
"5427","16.595570","cli","serv","IMAP","Request: AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA etc=="
"5433","16.596586","serv","cli","TCP","imap > 46434 [ACK] Seq=206 Ack=129 Win=5888 Len=0 TSV=3608644322 TSER=58648321"
"6420","18.341472","serv","cli","IMAP","Response: A1 OK Logged in."
"6421","18.341639","cli","serv","IMAP","Request: A2 CAPABILITY"
"6437","18.370916","serv","cli","TCP","imap > 46434 [ACK] Seq=224 Ack=144 Win=5888 Len=0 TSV=3608646052 TSER=58648757"
"6526","18.602063","serv","cli","IMAP","Response: * CAPABILITY IMAP4rev1 SASL-IR SORT THREAD=REFERENCES MULTIAPPEND UNSELECT LITERAL+ IDLE CHILDREN NAMESPACE LOGIN-REFERRALS QUOTA"
"6528","18.604615","cli","serv","IMAP","Request: A3 LIST "" inbox"
"6542","18.642682","serv","cli","TCP","imap > 46434 [ACK] Seq=384 Ack=162 Win=5888 Len=0 TSV=3608646316 TSER=58648823"
"6555","18.668482","serv","cli","IMAP","Response: * LIST (\HasNoChildren) "." "INBOX""
"6557","18.668945","cli","serv","IMAP","Request: A4 STATUS INBOX (MESSAGES RECENT UNSEEN UIDNEXT UIDVALIDITY)"
"6578","18.712320","serv","cli","IMAP","Response: * STATUS "INBOX" (MESSAGES 1 RECENT 0 UIDNEXT 10 UIDVALIDITY 1271690632 UNSEEN 1)"
"6584","18.713558","cli","serv","IMAP","Request: A5 LOGOUT"
"6588","18.728700","serv","cli","IMAP","Response: * BYE Logging out"
"6589","18.728774","serv","cli","IMAP","Response: A5 OK Logout completed."
"6590","18.728914","cli","serv","TCP","46434 > imap [ACK] Seq=235 Ack=597 Win=8000 Len=0 TSV=58648854 TSER=3608646398"
"6591","18.729118","cli","serv","TCP","46434 > imap [FIN, ACK] Seq=235 Ack=597 Win=8000 Len=0 TSV=58648854 TSER=3608646398"
It seems to me that the problem is server-side, in the login part.
Edited by: user13568923 on 18 janv. 2011 00:41

WAS JAVA 640 engine is giving performance problems.

Hi,
We are facing the performance issues in WAS java 640 SP19 engine. Pages are taking much time to load.
Kindly suggest if there anything do with jco connection parameter changes and recommended values. Is there any parametes to check at the backend system level.
WAS version: 640, SP19
OS : SUN 5.9
Oralce: 9.2.0.8
Thanks,
Hareesh

Hello Hareesh
You should find the root cause of your performance problem first.
Start with checking the server (cpu, memory, paging) and by checking if your portal runs on the recommended parameters (check oss notes).
A killer for portal performance is when you have a lot of paging going on, make sure the portal has enough extended memory available to avoid any paging.
If those don't give a result, you can still analyze garbage collection logs with third party tools.
Check if heapdumps are being generated and analyze them.
Once those checks are done, I would recommend installation of Sol Man Diagnostics + Wily Introscope so you get a better view on what is going on and troubleshoot further from there.
Kind regards
Tom

Messaging performance problem on RAID disk with many IMAP users

Similar Messages

Maybe you are looking for