antoine_mpo
Reputable Poster
Hi List,
Here is the context :
We use 2 platforms.
The first one is the development/integration platform : 1 web server, 1 enterprise server,1 deployment server, with DV7333, PY7333 and IN7333 environment.
The other one is outsourced, where we have preproduction and production environment. we have 1 preproduction web server, 1 entreprise (and database) server for preprod, 2 web servers for prod, 1 enterprise (and database) server for prod, 1 ube server for prod, 1 bsfn server for prod, 1 deployment server.
Here is the problem we have :
For months, preproduction is really hard to use because of unpredictable reponse time, in web or fat client. It's a real pain in the ass. For anything (search, order entry, ...) sometimes you get a normal response time, and the next time you can wait for minutes.
Here is an example : you want to search for all work orders of an order in P31125. So you type order number in RORN field of QBE, and click on search. Normally it should appears really quick (wich is the case in our DV, PY, IN or Prod environment), but sometimes (and it happens really often), you can wait for several seconds, up to 60 seconds ...
If you enter a simple order (we use configurator), which take something like 130-150 seconds from begining to end, it can take up to 800 seconds !!
Important note : preproduction is just used for some tests, so in full use, we are 2 or 3 connected max at a time !!
And it's not a WAN issue because we did measures during a week, 3 times/day, from here and from our supplier (where it's outsourced), on both web and fat client. We all get same unpredictable response time.
We did audit on database, jde.ini, ... nothing.
Servers are really big (they are supposed to replace production servers if needed), and they are really not stressed !!
Our supplier took a look at Oracle activity while using JDE, and when we wait in JDE, there's no activity in Oracle, so JDE is not waiting for big sql request answer. We don't know what it is waiting for !
I just did a test on fat client (located in our supplier building, we have remote access with VNC), connected in JPP7333 environment. I had a big wait time while searching for an order in P4210 (with criteria DOCO and DCTO), and in debugfile, i see a GetUserProfileCache that took 1069 seconds !
Does anyone have any ideas on what to check or where to look at ?
I can give you more information if you need so.
Thanks for your help.
Rgds,
Here is the context :
We use 2 platforms.
The first one is the development/integration platform : 1 web server, 1 enterprise server,1 deployment server, with DV7333, PY7333 and IN7333 environment.
The other one is outsourced, where we have preproduction and production environment. we have 1 preproduction web server, 1 entreprise (and database) server for preprod, 2 web servers for prod, 1 enterprise (and database) server for prod, 1 ube server for prod, 1 bsfn server for prod, 1 deployment server.
Here is the problem we have :
For months, preproduction is really hard to use because of unpredictable reponse time, in web or fat client. It's a real pain in the ass. For anything (search, order entry, ...) sometimes you get a normal response time, and the next time you can wait for minutes.
Here is an example : you want to search for all work orders of an order in P31125. So you type order number in RORN field of QBE, and click on search. Normally it should appears really quick (wich is the case in our DV, PY, IN or Prod environment), but sometimes (and it happens really often), you can wait for several seconds, up to 60 seconds ...
If you enter a simple order (we use configurator), which take something like 130-150 seconds from begining to end, it can take up to 800 seconds !!
Important note : preproduction is just used for some tests, so in full use, we are 2 or 3 connected max at a time !!
And it's not a WAN issue because we did measures during a week, 3 times/day, from here and from our supplier (where it's outsourced), on both web and fat client. We all get same unpredictable response time.
We did audit on database, jde.ini, ... nothing.
Servers are really big (they are supposed to replace production servers if needed), and they are really not stressed !!
Our supplier took a look at Oracle activity while using JDE, and when we wait in JDE, there's no activity in Oracle, so JDE is not waiting for big sql request answer. We don't know what it is waiting for !
I just did a test on fat client (located in our supplier building, we have remote access with VNC), connected in JPP7333 environment. I had a big wait time while searching for an order in P4210 (with criteria DOCO and DCTO), and in debugfile, i see a GetUserProfileCache that took 1069 seconds !
Does anyone have any ideas on what to check or where to look at ?
I can give you more information if you need so.
Thanks for your help.
Rgds,
