|
Hi, I'm in hurry and going out of office for 10 days, please send me another mail latter! some comments in the text On Monday 07 April 2003 22:27, you wrote: > Hi Dieter, > In the last week this is what we have done to find the possible bottle > neck in the system: > > We tried to run the application in isolation when almost the entire > system was at the disposal (9406-720, 2GB memory, 50% DASD utilization, > DZDASOINIT runs in machine pool etc...). This is what we have seen: > > 1. DB Monitor shows that actual SQL time is negligible. It takes about 1.8 > seconds to perform SQLs involved (about 25 SQLs). looks quite good We noticed that OPEN and > DE I assume direct execution (I don't know what is this -- description??) operations seem to take > most time out of this 1.8 seconds. Question is how do we make the opens > reuse ODPs? In our case every SQL seems to do full open. If an open of a cursor takes time, there might be an index missing. start the db monitor before making connection for all jobs and you will see debug messages. Note that we are > using JT400 to access AS400 data from servlets which are running on a PC > Webserver. Other question is what is this DE record and how do I reduce the > overhead there?? use prepared statements, wherever possible and cash already prepared statements for further use. > > 2. Overall response time is about 15-20 seconds. We feel that once the data > is obtained by servlet it should not take more than 1-2 seconds to create > the HTMl with images and render the page. normally milliseconds!!! Where else is the time lost?? Is > this thanx to JT400?? No, its not the driver (try more than 1 version and recent ones -> JTOPEN Note that the high speed PC server and AS400 are on > the LAN next to each other. > > Any suggestion would be helpful > Suresh > > > ----- Original Message ----- > From:Dieter.Bender@xxxxxxxxxxx (Dieter Bender) > To:Java Programming on and around the iSeries / AS400 > <java400-l@xxxxxxxxxxxx> > Subject:Re: SQL Insert > > > Hi Suresh, > > > > the test app runs on a pc as swing application, 10mb Ethernet, a little > > as400 > > 170, I'm using the toolbox driver. > > > > On Monday 31 March 2003 14:53, you wrote: > > > HI Dieter, > > > I am surprised to see the kind of performance you are getting out of > > > > your > > > > > AS400 box. Are u running the app natively on AS400? In our case, we > > > > are > > > > > usng JT400 for DB access and the java servlets run on a PC server > > under > > > IBM > > > > > WAS 4.0 app server. Both PC server and AS400 are on the TCP/IP LAN > > > > network > > > > > wth the speed 10/100 MBPS. Also, the AS400 is pretty high end > > > > (9406-720) > > > > > runnng V4R4 and is about 45% utilized. Plenty of disk space (80GB), > > > > memory > > > > > (2GB) and processor speed available. > > > > > > Do you think network config and/or JT400 is givign raise to this kind > > > > of > > > > > poor performance?? > > > > there must be a bottleneck, maybe main storage, maybe disk arms, > > maybe... > > the point to start - in my experience - is the database monitor and > > afterwards having alook at the paging rates of the pools, the disk i/o > > and > > the network traffic. > > starting the database monitor (it's part of the oops nav ...) with all > > details, I would have alook at the detailed single operations (3rd card > > left > > column first check box) they are sorted by time descending. second look > > at > > the summary of operations, a single database access has to be in the > > region > > of 1 millisecond in average. > > maybe somebody has done a little bit "optimization" on this box. I've > > seen > > boxes with 200 mb mainstorage for the database server jobs, trying to > > scratch > > holes in the 4 17 GB DASD. > > > > Dieter > > > > > Suresh > > > > > > ----- Original Message ----- > > > From: "Dieter Bender" <Dieter.Bender@xxxxxxxxxxx> > > > To: "Java Programming on and around the iSeries / AS400" > > > <java400-l@xxxxxxxxxxxx> > > > Sent: Thursday, March 27, 2003 2:46 PM > > > Subject: Re: SQL Insert > > > > > > > Hi Suresh, > > > > > > > > that seems quite a lot for me. On my little 170 it takes about 8 > > > > seconds > > > > > to > > > > > > > connect to the database and poll 10.000 records into a swing > > > > TableModel > > > > > > without any performance tuning, using plain JDBC or even SQLJ. Using > > > > a > > > > > real > > > > > > > databasee Server, this should be finished in 1 or 2 seconds, or even > > > > less. you should not invest your time in finding work arounds > > > > (stored > > > > > procedures, > > > > > > > record level access, ...) first thing should be to find the reason > > > > for > > > > > > the problem. what is the system doing all the time and wich > > > > component > > > > > > causes > > > > > > the > > > > > > > problem. > > > > Did you have a look with the Database Monitor to find out how long > > > > every > > > > > > request (Connect, open cursors, fetching records) to the database > > > > took??? > > > > > > Did you have a look to the overall workload (CPU, memory, disk > > > > activity, > > > > > > network traffic) of the system with management central to find the > > > > bottleneck??? > > > > > > > > Dieter > > > > > > > > On Thursday 27 March 2003 06:53, you wrote: > > > > > Hi Dieter, > > > > > This is the statistics: > > > > > 1. We do about 25 SQL statements against 4 tables. > > > > > > > > > > 2. All the tables are small (about 100 records/table) except one > > > > which > > > > > has > > > > > > > > about 10,000 records. > > > > > > > > > > 3. Ultimately we end up getting about 60-70 records from these > > > > tables > > > > > which > > > > > > > > is needed for processing by the servlet (at the App server). > > > > > > > > > > 4. I have 2 questions in connection with this: > > > > > a. In the stored procedure (written in RPGLE), we are basically > > > > > peforming various SQLs and returning the result set cursor back to > > > > the > > > > > Java > > > > > > > > client program. This RPG program declares the cursors (SQLs), open > > > > the > > > > > > > cursors and return them: all together about 15 result sets. Other > > > > > option > > > > > > is > > > > > > > > to call the RPG program repeatedly so that it will send back > > > > proper > > > > > cursor > > > > > > > > one at a time. Which will be faster?? > > > > > > > > > > b. Also, is it faster with array result sets as opposed to > > > > cursor > > > > > result > > > > > > > > sets?? How can the client read an array result set in the Java > > > > servlet? > > > > > > > Suresh > > > > > Antares Systems Ltd. > > > > > # 24, Sudha Complex, 3rd Stage, 4th Block, > > > > > Basaveshwaranagar, > > > > > Bangalore - 560 079 > > > > > Telefax : 91-80-3284311, 3284313 > > > > > _______________________________________________ > > > > > This is the Java Programming on and around the iSeries / AS400 > > > > > > (JAVA400-L) > > > > > > > > mailing list To post a message email: JAVA400-L@xxxxxxxxxxxx > > > > > To subscribe, unsubscribe, or change list options, > > > > > visit: http://lists.midrange.com/mailman/listinfo/java400-l > > > > > or email: JAVA400-L-request@xxxxxxxxxxxx > > > > > Before posting, please take a moment to review the archives > > > > > at http://archive.midrange.com/java400-l. > > > > > > > > -- > > > > mfG > > > > > > > > Dieter Bender > > > > > > > > > > > > DV-Beratung Dieter Bender > > > > Wetzlarerstr. 25 > > > > 35435 Wettenberg > > > > Tel. +49 641 9805855 > > > > Fax +49 641 9805856 > > > > www.bender-dv.de > > > > _______________________________________________ > > > > This is the Java Programming on and around the iSeries / AS400 > > > > (JAVA400-L) > > > > > > mailing list > > > > > > > To post a message email: JAVA400-L@xxxxxxxxxxxx > > > > To subscribe, unsubscribe, or change list options, > > > > visit: http://lists.midrange.com/mailman/listinfo/java400-l > > > > or email: JAVA400-L-request@xxxxxxxxxxxx > > > > Before posting, please take a moment to review the archives > > > > at http://archive.midrange.com/java400-l. > > > > > > Antares Systems Ltd. > > > # 24, Sudha Complex, 3rd Stage, 4th Block, > > > Basaveshwaranagar, > > > Bangalore - 560 079 > > > Telefax : 91-80-3284311, 3284313 > > > _______________________________________________ > > > This is the Java Programming on and around the iSeries / AS400 > > > > (JAVA400-L) > > > > > mailing list To post a message email: JAVA400-L@xxxxxxxxxxxx > > > To subscribe, unsubscribe, or change list options, > > > visit: http://lists.midrange.com/mailman/listinfo/java400-l > > > or email: JAVA400-L-request@xxxxxxxxxxxx > > > Before posting, please take a moment to review the archives > > > at http://archive.midrange.com/java400-l. > > > > -- > > mfG > > > > Dieter Bender > > > > > > DV-Beratung Dieter Bender > > Wetzlarerstr. 25 > > 35435 Wettenberg > > Tel. +49 641 9805855 > > Fax +49 641 9805856 > > www.bender-dv.de > > _______________________________________________ > > This is the Java Programming on and around the iSeries / AS400 > > (JAVA400-L) mailing list > > To post a message email: JAVA400-L@xxxxxxxxxxxx > > To subscribe, unsubscribe, or change list options, > > visit: http://lists.midrange.com/mailman/listinfo/java400-l > > or email: JAVA400-L-request@xxxxxxxxxxxx > > Before posting, please take a moment to review the archives > > at http://archive.midrange.com/java400-l. > Antares Systems Ltd. > # 24, Sudha Complex, 3rd Stage, 4th Block, > Basaveshwaranagar, > Bangalore - 560 079 > Telefax : 91-80-3284311, 3284313 > _______________________________________________ > This is the Java Programming on and around the iSeries / AS400 (JAVA400-L) > mailing list To post a message email: JAVA400-L@xxxxxxxxxxxx > To subscribe, unsubscribe, or change list options, > visit: http://lists.midrange.com/mailman/listinfo/java400-l > or email: JAVA400-L-request@xxxxxxxxxxxx > Before posting, please take a moment to review the archives > at http://archive.midrange.com/java400-l. -- mfG Dieter Bender DV-Beratung Dieter Bender Wetzlarerstr. 25 35435 Wettenberg Tel. +49 641 9805855 Fax +49 641 9805856 www.bender-dv.de
As an Amazon Associate we earn from qualifying purchases.
This mailing list archive is Copyright 1997-2024 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].
Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.