Re: Mid-tier prefetch scripts

2007-01-19 Thread Tony Worthington
I second the Tomcat performance vote.  It's noticably faster (even using 
Apache and mod_jk) than IIS/SE(AS) even without using precache/fetch.  And 
compared to versions 5/6 it flies.

Chris - you don't mention what environment you're operating in?

Our web boxes are Win2k3 2x3.8GHz dual core xeon with 3GB of ram.

-- 
Tony Worthington
[EMAIL PROTECTED]
262-703-5911



Axton [EMAIL PROTECTED] 
Sent by: Action Request System discussion list(ARSList) 
arslist@ARSLIST.ORG
01/19/2007 08:17 AM
Please respond to
arslist@ARSLIST.ORG


To
arslist@ARSLIST.ORG
cc

Subject
Re: Mid-tier prefetch scripts






** 
I was happy with the mid-tier performance using jboss or tomcat with the 
latest 1.5 jdk, but people here complain about performance when using IBM 
WebSphere 6 with the IBM 1.4 jdk.
 
Axton

 
On 1/19/07, Chris Akens [EMAIL PROTECTED] wrote: 
From what I understand, the mid-tier pre-caching is actually done on a
unique permission group list basis. Therefore you would want to pre-cache 
with an account that has the same permissions as the majority of your 
users.
I have not used the pre-cache functionality that is included as part of 
7.0;
but am interested in how it takes into account the different permission 
group combinations. Any insight into this matter would be appreciated.

I have been less than impressed with the mid-tier v7 thus far to say the
least. Any suggestions on tuning the mid-tier for performance?

___
UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:Where 
the Answers Are 

__20060125___This posting was submitted with HTML in 
it___ 


CONFIDENTIALITY NOTICE: 
This is a transmission from Kohl's Department Stores, Inc.
and may contain information which is confidential and proprietary.
If you are not the addressee, any disclosure, copying or distribution or use of 
the contents of this message is expressly prohibited.
If you have received this transmission in error, please destroy it and notify 
us immediately at 262-703-7000.

CAUTION:
Internet and e-mail communications are Kohl's property and Kohl's reserves the 
right to retrieve and read any message created, sent and received.  Kohl's 
reserves the right to monitor messages to or from authorized Kohl's Associates 
at any time
without any further consent.

___
UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:Where the 
Answers Are


Re: Mid-tier prefetch scripts

2007-01-19 Thread Carey Matthew Black

Chris also did not define what less than impressed means either.

What kind of performance are you seeing that is not impressing you?
(sub 60 second form open? sub 10 second form open?  All forms or just
some forms?)

--
Carey Matthew Black
Remedy Skilled Professional (RSP)
ARS = Action Request System(Remedy)

Love, then teach
Solution = People + Process + Tools
Fast, Accurate, Cheap Pick two.

On 1/19/07, Tony Worthington [EMAIL PROTECTED] wrote:

I second the Tomcat performance vote.  It's noticably faster (even using
Apache and mod_jk) than IIS/SE(AS) even without using precache/fetch.  And
compared to versions 5/6 it flies.

Chris - you don't mention what environment you're operating in?

Our web boxes are Win2k3 2x3.8GHz dual core xeon with 3GB of ram.

--
Tony Worthington
[EMAIL PROTECTED]
262-703-5911


___
UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:Where the Answers 
Are


Re: Mid-tier prefetch scripts

2007-01-19 Thread strauss
My response times were measured on both the production server set that I
just built out, and development set which is identical except for
mid-tier:

Web - Win2K3 R2 Enterprise x64 on HP DL380 G5 with two 4-core Xeons and
12 gb RAM
   Mid-tier 7.0.01 on Tomcat app server AND Tomcat web, JDK 1.4.2_13 -
Tomcat has memory pool from 512 to 1536 defined
   The development mid-tier is identical except that the server is
Win2K3 Ent x86 on a DL-380 with two 3.4 ghz Xeons and 6 gb RAM

ARS - Win2K3 Enterprise x64 on HP DL385 with two 2-core Opterons and 10
gb RAM
DB - Win2K3 Enterprise x64 and SQL Server 2005 x64 on HP DL385 with two
2-core Opterons and 10 gb RAM (SQL Server grabs 8 gb RAM)
Client - WinXP Pro and IE 7 on 3ghz PC with 1 gb RAM

The long lag times to load non-prefetched forms (1:30 for Change
console, 1:40 for New Incident opened from Overview console) tells me
that basic mid-tier 7 architecture still leaves a lot to be desired.
Almost all of the sustained CPU load when pre-fetching objects and
loading a form for the first time is on the AR server, which sits at
30-35% total CPU for up to 1 to 1.5 minutes in the worst cases; one CPU
will stay at 75%, one at 50%, one at 25%. The CPU traces on the mid-tier
servers peak at 20% intermittently on several processors - usually 2 of
them. There is usually a single grab from the db server - one spike on
one CPU, or no trace at all in some cases!

Less than happy??? Any time you perform an operation for the first time,
like search for a customer from a new ticket, you will wish you were
back on ITSM 5.5 and mid-tier 5.1.2 when it comes to performance. It is
not a user experience that I want my IT staff to have to live with - who
do you know that will consider it acceptable to click on a link to open
a console or form and then stare at a blank, white screen for 1 and a
half minutes? Nobody here! All response times _after_ forms have loaded
once, either prefetched or not, were very fast - in the 1 to 2 seconds
range. It's that first impression that is going to be a killer. It
even occurs on little dialog boxes, like the selection box to Create a
request from the Overview console, or search for a customer. Each new
form or dialog that you invoke for the first time has a long, pregnant
pause before it displays - especially if it was not included in a
prefetch.

Christopher Strauss, Ph.D.
Remedy Database Administrator
University of North Texas Computing Center
http://remedy.unt.edu/helpdesk/
-Original Message-
From: Action Request System discussion list(ARSList)
[mailto:[EMAIL PROTECTED] On Behalf Of Tony Worthington
Sent: Friday, January 19, 2007 8:28 AM
To: arslist@ARSLIST.ORG
Subject: Re: Mid-tier prefetch scripts

I second the Tomcat performance vote.  It's noticably faster (even using
Apache and mod_jk) than IIS/SE(AS) even without using precache/fetch.
And compared to versions 5/6 it flies.

Chris - you don't mention what environment you're operating in?

Our web boxes are Win2k3 2x3.8GHz dual core xeon with 3GB of ram.

--
Tony Worthington
[EMAIL PROTECTED]
262-703-5911



Axton [EMAIL PROTECTED] 
Sent by: Action Request System discussion list(ARSList) 
arslist@ARSLIST.ORG
01/19/2007 08:17 AM
Please respond to
arslist@ARSLIST.ORG


To
arslist@ARSLIST.ORG
cc

Subject
Re: Mid-tier prefetch scripts






** 
I was happy with the mid-tier performance using jboss or tomcat with the

latest 1.5 jdk, but people here complain about performance when using
IBM 
WebSphere 6 with the IBM 1.4 jdk.
 
Axton

 
On 1/19/07, Chris Akens [EMAIL PROTECTED] wrote: 
From what I understand, the mid-tier pre-caching is actually done on a
unique permission group list basis. Therefore you would want to
pre-cache 
with an account that has the same permissions as the majority of your 
users.
I have not used the pre-cache functionality that is included as part of 
7.0;
but am interested in how it takes into account the different permission 
group combinations. Any insight into this matter would be appreciated.

I have been less than impressed with the mid-tier v7 thus far to say the
least. Any suggestions on tuning the mid-tier for performance?


___
UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:Where

the Answers Are 

__20060125___This posting was submitted with HTML in

it___ 


CONFIDENTIALITY NOTICE: 
This is a transmission from Kohl's Department Stores, Inc.
and may contain information which is confidential and proprietary.
If you are not the addressee, any disclosure, copying or distribution or
use of the contents of this message is expressly prohibited.
If you have received this transmission in error, please destroy it and
notify us immediately at 262-703-7000.

CAUTION:
Internet and e-mail communications are Kohl's property and Kohl's
reserves the right to retrieve and read any message created, sent and
received.  Kohl's reserves the right to monitor

Re: Mid-tier prefetch scripts

2007-01-19 Thread Tony Worthington
I don't disagree that the initial rendering is quite slow.  And it sucks 
that the only solution is to either prefetch via a perl/whatever script, 
or use the xml midtier method.

My point was more towards post-initial-form/view-load (cached) 
performance...

Now if there was a script that dumped every web viewable form into that 
prefetch xml script... (we reboot weekly)

 :-)


-- 
Tony Worthington
[EMAIL PROTECTED]
262-703-5911



strauss [EMAIL PROTECTED] 
Sent by: Action Request System discussion list(ARSList) 
arslist@ARSLIST.ORG
01/19/2007 12:03 PM
Please respond to
arslist@ARSLIST.ORG


To
arslist@ARSLIST.ORG
cc

Subject
Re: Mid-tier prefetch scripts






My response times were measured on both the production server set that I
just built out, and development set which is identical except for
mid-tier:

Web - Win2K3 R2 Enterprise x64 on HP DL380 G5 with two 4-core Xeons and
12 gb RAM
   Mid-tier 7.0.01 on Tomcat app server AND Tomcat web, JDK 1.4.2_13 -
Tomcat has memory pool from 512 to 1536 defined
   The development mid-tier is identical except that the server is
Win2K3 Ent x86 on a DL-380 with two 3.4 ghz Xeons and 6 gb RAM

ARS - Win2K3 Enterprise x64 on HP DL385 with two 2-core Opterons and 10
gb RAM
DB - Win2K3 Enterprise x64 and SQL Server 2005 x64 on HP DL385 with two
2-core Opterons and 10 gb RAM (SQL Server grabs 8 gb RAM)
Client - WinXP Pro and IE 7 on 3ghz PC with 1 gb RAM

The long lag times to load non-prefetched forms (1:30 for Change
console, 1:40 for New Incident opened from Overview console) tells me
that basic mid-tier 7 architecture still leaves a lot to be desired.
Almost all of the sustained CPU load when pre-fetching objects and
loading a form for the first time is on the AR server, which sits at
30-35% total CPU for up to 1 to 1.5 minutes in the worst cases; one CPU
will stay at 75%, one at 50%, one at 25%. The CPU traces on the mid-tier
servers peak at 20% intermittently on several processors - usually 2 of
them. There is usually a single grab from the db server - one spike on
one CPU, or no trace at all in some cases!

Less than happy??? Any time you perform an operation for the first time,
like search for a customer from a new ticket, you will wish you were
back on ITSM 5.5 and mid-tier 5.1.2 when it comes to performance. It is
not a user experience that I want my IT staff to have to live with - who
do you know that will consider it acceptable to click on a link to open
a console or form and then stare at a blank, white screen for 1 and a
half minutes? Nobody here! All response times _after_ forms have loaded
once, either prefetched or not, were very fast - in the 1 to 2 seconds
range. It's that first impression that is going to be a killer. It
even occurs on little dialog boxes, like the selection box to Create a
request from the Overview console, or search for a customer. Each new
form or dialog that you invoke for the first time has a long, pregnant
pause before it displays - especially if it was not included in a
prefetch.

Christopher Strauss, Ph.D.
Remedy Database Administrator
University of North Texas Computing Center
http://remedy.unt.edu/helpdesk/
-Original Message-
From: Action Request System discussion list(ARSList)
[mailto:[EMAIL PROTECTED] On Behalf Of Tony Worthington
Sent: Friday, January 19, 2007 8:28 AM
To: arslist@ARSLIST.ORG
Subject: Re: Mid-tier prefetch scripts

I second the Tomcat performance vote.  It's noticably faster (even using
Apache and mod_jk) than IIS/SE(AS) even without using precache/fetch.
And compared to versions 5/6 it flies.

Chris - you don't mention what environment you're operating in?

Our web boxes are Win2k3 2x3.8GHz dual core xeon with 3GB of ram.

--
Tony Worthington
[EMAIL PROTECTED]
262-703-5911



Axton [EMAIL PROTECTED] 
Sent by: Action Request System discussion list(ARSList) 
arslist@ARSLIST.ORG
01/19/2007 08:17 AM
Please respond to
arslist@ARSLIST.ORG


To
arslist@ARSLIST.ORG
cc

Subject
Re: Mid-tier prefetch scripts






** 
I was happy with the mid-tier performance using jboss or tomcat with the

latest 1.5 jdk, but people here complain about performance when using
IBM 
WebSphere 6 with the IBM 1.4 jdk.
 
Axton

 
On 1/19/07, Chris Akens [EMAIL PROTECTED] wrote: 
From what I understand, the mid-tier pre-caching is actually done on a
unique permission group list basis. Therefore you would want to
pre-cache 
with an account that has the same permissions as the majority of your 
users.
I have not used the pre-cache functionality that is included as part of 
7.0;
but am interested in how it takes into account the different permission 
group combinations. Any insight into this matter would be appreciated.

I have been less than impressed with the mid-tier v7 thus far to say the
least. Any suggestions on tuning the mid-tier for performance?


___
UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:Where

the Answers

Mid-tier prefetch scripts

2007-01-18 Thread strauss
Does anyone think that they have figured out all of the ins and outs of
using the prefetch script with mid-tier (prefetchConfig.xml)?? Since
this is a largely undocumented utility that BMC Remedy Support does not
fully understand, I'll take any hints that others have already figured
out. The latest script I am trying to use for ITSM 7.0 looks like this:

?xml version=1.0 encoding=UTF-8?
midtier-prefetch-config  xmlns=http://www.bmc.com/remedy/midtier/700;
  prefetch-user
user-nameDemo/user-name
localeen_US/locale
prefetch-server
  server-nameremedy6.ars.unt.edu/server-name
prefetch-app
  app-nameRemedy Incident Management/app-name
/prefetch-app
prefetch-app
  app-nameRemedy Change Management/app-name
/prefetch-app
prefetch-app
  app-nameRemedy Change Management Dashboard/app-name
/prefetch-app
prefetch-app
  app-nameRemedy Problem Management/app-name
/prefetch-app
prefetch-app
  app-nameRemedy Requester Console/app-name
/prefetch-app
prefetch-app
  app-nameRemedy Task Management/app-name
/prefetch-app
prefetch-app
  app-nameRemedy Reporting Console/app-name
/prefetch-app
/prefetch-server
prefetch-server
  server-nameremedy6.ars.unt.edu/server-name
  prefetch-app
  app-nameRemedy Foundation Elements/app-name
prefetch-form
   form-nameSHR:OverviewConsole/form-name
/prefetch-form
/prefetch-app
/prefetch-server
prefetch-server
  server-nameremedy6.ars.unt.edu/server-name
prefetch-form
   form-nameHome Page/form-name
/prefetch-form
prefetch-form
   form-nameSHR:OverviewConsole/form-name
/prefetch-form
prefetch-form
   form-nameHPD:Help Desk/form-name
/prefetch-form
prefetch-form
   form-nameCHG:Infrastructure Change/form-name
/prefetch-form
prefetch-form
   form-namePBM:Known Error/form-name
/prefetch-form
prefetch-form
   form-namePBM:Problem Investigation/form-name
/prefetch-form
prefetch-form
   form-namePBM:Solution Database/form-name
/prefetch-form
prefetch-form
   form-nameTMS:Task/form-name
/prefetch-form
/prefetch-server
  /prefetch-user
/midtier-prefetch-config

I am not convinced that everything in the script is in fact prefetching
yet. It takes over 30 minutes for prefetch to run after the mid-tier
server starts (the mid-tier shows little CPU usage but the AR Server
sits at 30% utilization throughout!), but the difference in performance
for the first time a form loads can be huge. Also, that difference
varies based on who is logging in - the script was run with Demo, tests
were with a support staff login (AREA authenticated) and then with the
appadmin account, which was slower:

Console IncidentChange Problem   Overview
Requester

Not Prefetched load time  0:39   1:35   0:25   0:47
0:26
Prefetched, 1st time load 0:04   0:22   0:04   0:27
0:08
(same, but different user)0:16   0:53   0:12   0:08
0:29
All subsequent loadings   0:02   0:02   0:02   0:02
0:01

Loading forms from the consoles also differs depending on whether the
form is part of the same application that the console is in:

FormNew Incident from Incident Console   New
Incident from Overview Console

Not Prefetched load time0:24
1:40  
Prefetched, 1st time load   0:06
0:04   
All subsequent loadings 0:03
0:03

I now believe that this utility is an essential part of any
mid-tier/ITSM 7.x installation, since the load times measured for
initial ITSM 7 form access are clearly unacceptable. Prefetch will need
to be a lot better documented and understood before that will be
possible, so please share any experience that you might have with it.

Christopher Strauss, Ph.D.
Remedy Database Administrator
University of North Texas Computing Center
http://remedy.unt.edu/helpdesk/

___
UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:Where the 
Answers Are


Re: Mid-tier prefetch scripts

2007-01-18 Thread Kyle Whitley

Is this only for ARS/Mid Tier 7.x  and ITSM 7.x ?


The only prefetch script I have seen is a perl script  not anything in 
xml.  I will say that with the perl script and my testing(ITSM 6 and ARS 
6.3), load times  really depended on privileges of the user that was 
used to authenticate to the AR System in the script.  Meaning if I used 
an Admin login in the script the admin views loaded quicker than a view 
a support person would receive on initial load of the view.  I 
experienced similar load times using the perl script in ARS 6.3, but my 
runtime of the perl script was shorter,  I wasn't loading as many forms 
as you have.  I only ran the script on requester views for self service.


Kyle

strauss wrote:

Does anyone think that they have figured out all of the ins and outs of
using the prefetch script with mid-tier (prefetchConfig.xml)?? Since
this is a largely undocumented utility that BMC Remedy Support does not
fully understand, I'll take any hints that others have already figured
out. The latest script I am trying to use for ITSM 7.0 looks like this:

?xml version=1.0 encoding=UTF-8?
midtier-prefetch-config  xmlns=http://www.bmc.com/remedy/midtier/700;
  prefetch-user
user-nameDemo/user-name
localeen_US/locale
prefetch-server
  server-nameremedy6.ars.unt.edu/server-name
prefetch-app
  app-nameRemedy Incident Management/app-name
/prefetch-app
prefetch-app
  app-nameRemedy Change Management/app-name
/prefetch-app
prefetch-app
  app-nameRemedy Change Management Dashboard/app-name
/prefetch-app
prefetch-app
  app-nameRemedy Problem Management/app-name
/prefetch-app
prefetch-app
  app-nameRemedy Requester Console/app-name
/prefetch-app
prefetch-app
  app-nameRemedy Task Management/app-name
/prefetch-app
prefetch-app
  app-nameRemedy Reporting Console/app-name
/prefetch-app
/prefetch-server
prefetch-server
  server-nameremedy6.ars.unt.edu/server-name
  prefetch-app
  app-nameRemedy Foundation Elements/app-name
prefetch-form
   form-nameSHR:OverviewConsole/form-name
/prefetch-form
/prefetch-app
/prefetch-server
prefetch-server
  server-nameremedy6.ars.unt.edu/server-name
prefetch-form
   form-nameHome Page/form-name
/prefetch-form
prefetch-form
   form-nameSHR:OverviewConsole/form-name
/prefetch-form
prefetch-form
   form-nameHPD:Help Desk/form-name
/prefetch-form
prefetch-form
   form-nameCHG:Infrastructure Change/form-name
/prefetch-form
prefetch-form
   form-namePBM:Known Error/form-name
/prefetch-form
prefetch-form
   form-namePBM:Problem Investigation/form-name
/prefetch-form
prefetch-form
   form-namePBM:Solution Database/form-name
/prefetch-form
prefetch-form
   form-nameTMS:Task/form-name
/prefetch-form
/prefetch-server
  /prefetch-user
/midtier-prefetch-config

I am not convinced that everything in the script is in fact prefetching
yet. It takes over 30 minutes for prefetch to run after the mid-tier
server starts (the mid-tier shows little CPU usage but the AR Server
sits at 30% utilization throughout!), but the difference in performance
for the first time a form loads can be huge. Also, that difference
varies based on who is logging in - the script was run with Demo, tests
were with a support staff login (AREA authenticated) and then with the
appadmin account, which was slower:

Console IncidentChange Problem   Overview
Requester

Not Prefetched load time  0:39   1:35   0:25   0:47
0:26
Prefetched, 1st time load 0:04   0:22   0:04   0:27
0:08
(same, but different user)0:16   0:53   0:12   0:08
0:29
All subsequent loadings   0:02   0:02   0:02   0:02
0:01

Loading forms from the consoles also differs depending on whether the
form is part of the same application that the console is in:

FormNew Incident from Incident Console   New
Incident from Overview Console

Not Prefetched load time0:24
1:40  
Prefetched, 1st time load   0:06
0:04   
All subsequent loadings 0:03

0:03

I now believe that this utility is an essential part of any
mid-tier/ITSM 7.x installation, since the load times measured for
initial ITSM 7 form access are clearly unacceptable. Prefetch will need
to be a lot better documented and understood before that will be
possible, so please share any experience that you might have with it.

Christopher Strauss, Ph.D.
Remedy Database Administrator
University of North Texas Computing Center

Re: Mid-tier prefetch scripts

2007-01-18 Thread Mike Gould

I was also only aware of the Perl prefetch script.  Didn't realize there was
a new XML-based one for Mid-Tier 7.x.  Is this available for download
somewhere or do I need to contact Support?

I have to say that the Perl script never worked very well and like Kyle
said, seemed to only speed up load times for the user that is was executed
by but for no one else.  It sounds like that has been largely corrected but
not completely.

Every little bit helps though!

Thanks,
Michael Gould
US House of Representatives


On 1/18/07, Kyle Whitley [EMAIL PROTECTED] wrote:


Is this only for ARS/Mid Tier 7.x  and ITSM 7.x ?


The only prefetch script I have seen is a perl script  not anything in
xml.  I will say that with the perl script and my testing(ITSM 6 and ARS
6.3), load times  really depended on privileges of the user that was
used to authenticate to the AR System in the script.  Meaning if I used
an Admin login in the script the admin views loaded quicker than a view
a support person would receive on initial load of the view.  I
experienced similar load times using the perl script in ARS 6.3, but my
runtime of the perl script was shorter,  I wasn't loading as many forms
as you have.  I only ran the script on requester views for self service.

Kyle

strauss wrote:
 Does anyone think that they have figured out all of the ins and outs of
 using the prefetch script with mid-tier (prefetchConfig.xml)?? Since
 this is a largely undocumented utility that BMC Remedy Support does not
 fully understand, I'll take any hints that others have already figured
 out. The latest script I am trying to use for ITSM 7.0 looks like this:

 ?xml version=1.0 encoding=UTF-8?
 midtier-prefetch-config  xmlns=http://www.bmc.com/remedy/midtier/700;
   prefetch-user
 user-nameDemo/user-name
 localeen_US/locale
 prefetch-server
   server-nameremedy6.ars.unt.edu/server-name
 prefetch-app
   app-nameRemedy Incident Management/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Change Management/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Change Management Dashboard/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Problem Management/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Requester Console/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Task Management/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Reporting Console/app-name
 /prefetch-app
 /prefetch-server
 prefetch-server
   server-nameremedy6.ars.unt.edu/server-name
   prefetch-app
   app-nameRemedy Foundation Elements/app-name
 prefetch-form
form-nameSHR:OverviewConsole/form-name
 /prefetch-form
 /prefetch-app
 /prefetch-server
 prefetch-server
   server-nameremedy6.ars.unt.edu/server-name
 prefetch-form
form-nameHome Page/form-name
 /prefetch-form
 prefetch-form
form-nameSHR:OverviewConsole/form-name
 /prefetch-form
 prefetch-form
form-nameHPD:Help Desk/form-name
 /prefetch-form
 prefetch-form
form-nameCHG:Infrastructure Change/form-name
 /prefetch-form
 prefetch-form
form-namePBM:Known Error/form-name
 /prefetch-form
 prefetch-form
form-namePBM:Problem Investigation/form-name
 /prefetch-form
 prefetch-form
form-namePBM:Solution Database/form-name
 /prefetch-form
 prefetch-form
form-nameTMS:Task/form-name
 /prefetch-form
 /prefetch-server
   /prefetch-user
 /midtier-prefetch-config

 I am not convinced that everything in the script is in fact prefetching
 yet. It takes over 30 minutes for prefetch to run after the mid-tier
 server starts (the mid-tier shows little CPU usage but the AR Server
 sits at 30% utilization throughout!), but the difference in performance
 for the first time a form loads can be huge. Also, that difference
 varies based on who is logging in - the script was run with Demo, tests
 were with a support staff login (AREA authenticated) and then with the
 appadmin account, which was slower:

 Console IncidentChange Problem   Overview
 Requester

 Not Prefetched load time  0:39   1:35   0:25   0:47
 0:26
 Prefetched, 1st time load 0:04   0:22   0:04   0:27
 0:08
 (same, but different user)0:16   0:53   0:12   0:08
 0:29
 All subsequent loadings   0:02   0:02   0:02   0:02
 0:01

 Loading forms from the consoles also differs depending on whether the
 form is part of the same application that the console is in:

 Form

Re: Mid-tier prefetch scripts

2007-01-18 Thread Axton

I believe the prefetch.xml is driven by the permissions of the account used
to perform the prefetch.  That's to say that using an admin account to
perform the prefetch will not have as much of a benefit for non-permission
users accessing the system.  I configured our servers to use a no-permission
account to perform the prefetch, since that is the permission base of our
end users that access the mid-tier interface.

Axton Grams


On 1/18/07, Kyle Whitley [EMAIL PROTECTED] wrote:


Is this only for ARS/Mid Tier 7.x  and ITSM 7.x ?


The only prefetch script I have seen is a perl script  not anything in
xml.  I will say that with the perl script and my testing(ITSM 6 and ARS
6.3), load times  really depended on privileges of the user that was
used to authenticate to the AR System in the script.  Meaning if I used
an Admin login in the script the admin views loaded quicker than a view
a support person would receive on initial load of the view.  I
experienced similar load times using the perl script in ARS 6.3, but my
runtime of the perl script was shorter,  I wasn't loading as many forms
as you have.  I only ran the script on requester views for self service.

Kyle

strauss wrote:
 Does anyone think that they have figured out all of the ins and outs of
 using the prefetch script with mid-tier (prefetchConfig.xml)?? Since
 this is a largely undocumented utility that BMC Remedy Support does not
 fully understand, I'll take any hints that others have already figured
 out. The latest script I am trying to use for ITSM 7.0 looks like this:

 ?xml version=1.0 encoding=UTF-8?
 midtier-prefetch-config  xmlns=http://www.bmc.com/remedy/midtier/700;
   prefetch-user
 user-nameDemo/user-name
 localeen_US/locale
 prefetch-server
   server-nameremedy6.ars.unt.edu/server-name
 prefetch-app
   app-nameRemedy Incident Management/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Change Management/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Change Management Dashboard/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Problem Management/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Requester Console/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Task Management/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Reporting Console/app-name
 /prefetch-app
 /prefetch-server
 prefetch-server
   server-nameremedy6.ars.unt.edu/server-name
   prefetch-app
   app-nameRemedy Foundation Elements/app-name
 prefetch-form
form-nameSHR:OverviewConsole/form-name
 /prefetch-form
 /prefetch-app
 /prefetch-server
 prefetch-server
   server-nameremedy6.ars.unt.edu/server-name
 prefetch-form
form-nameHome Page/form-name
 /prefetch-form
 prefetch-form
form-nameSHR:OverviewConsole/form-name
 /prefetch-form
 prefetch-form
form-nameHPD:Help Desk/form-name
 /prefetch-form
 prefetch-form
form-nameCHG:Infrastructure Change/form-name
 /prefetch-form
 prefetch-form
form-namePBM:Known Error/form-name
 /prefetch-form
 prefetch-form
form-namePBM:Problem Investigation/form-name
 /prefetch-form
 prefetch-form
form-namePBM:Solution Database/form-name
 /prefetch-form
 prefetch-form
form-nameTMS:Task/form-name
 /prefetch-form
 /prefetch-server
   /prefetch-user
 /midtier-prefetch-config

 I am not convinced that everything in the script is in fact prefetching
 yet. It takes over 30 minutes for prefetch to run after the mid-tier
 server starts (the mid-tier shows little CPU usage but the AR Server
 sits at 30% utilization throughout!), but the difference in performance
 for the first time a form loads can be huge. Also, that difference
 varies based on who is logging in - the script was run with Demo, tests
 were with a support staff login (AREA authenticated) and then with the
 appadmin account, which was slower:

 Console IncidentChange Problem   Overview
 Requester

 Not Prefetched load time  0:39   1:35   0:25   0:47
 0:26
 Prefetched, 1st time load 0:04   0:22   0:04   0:27
 0:08
 (same, but different user)0:16   0:53   0:12   0:08
 0:29
 All subsequent loadings   0:02   0:02   0:02   0:02
 0:01

 Loading forms from the consoles also differs depending on whether the
 form is part of the same application that the console is in:

 FormNew Incident from Incident Console   New
 Incident from Overview Console

 Not 

Re: Mid-tier prefetch scripts

2007-01-18 Thread Axton

It's included with mid-tier 7.x

Axton Grams


On 1/18/07, Mike Gould [EMAIL PROTECTED] wrote:


** I was also only aware of the Perl prefetch script.  Didn't realize
there was a new XML-based one for Mid-Tier 7.x.  Is this available for
download somewhere or do I need to contact Support?

I have to say that the Perl script never worked very well and like Kyle
said, seemed to only speed up load times for the user that is was executed
by but for no one else.  It sounds like that has been largely corrected but
not completely.

Every little bit helps though!

Thanks,
Michael Gould
US House of Representatives


On 1/18/07, Kyle Whitley [EMAIL PROTECTED] wrote:

 Is this only for ARS/Mid Tier 7.x  and ITSM 7.x ?


 The only prefetch script I have seen is a perl script  not anything in
 xml.  I will say that with the perl script and my testing(ITSM 6 and ARS
 6.3), load times  really depended on privileges of the user that was
 used to authenticate to the AR System in the script.  Meaning if I used
 an Admin login in the script the admin views loaded quicker than a view
 a support person would receive on initial load of the view.  I
 experienced similar load times using the perl script in ARS 6.3, but my
 runtime of the perl script was shorter,  I wasn't loading as many forms
 as you have.  I only ran the script on requester views for self service.

 Kyle

 strauss wrote:
  Does anyone think that they have figured out all of the ins and outs
 of
  using the prefetch script with mid-tier ( prefetchConfig.xml)?? Since
  this is a largely undocumented utility that BMC Remedy Support does
 not
  fully understand, I'll take any hints that others have already figured
  out. The latest script I am trying to use for ITSM 7.0 looks like
 this:
 
  ?xml version=1.0 encoding=UTF-8?
  midtier-prefetch-config  xmlns=http://www.bmc.com/remedy/midtier/700
 
prefetch-user
  user-nameDemo/user-name
  localeen_US/locale
  prefetch-server
server-name remedy6.ars.unt.edu/server-name
  prefetch-app
app-nameRemedy Incident Management/app-name
  /prefetch-app
  prefetch-app
app-nameRemedy Change Management/app-name
  /prefetch-app
  prefetch-app
app-nameRemedy Change Management Dashboard/app-name
  /prefetch-app
  prefetch-app
app-nameRemedy Problem Management/app-name
  /prefetch-app
  prefetch-app
app-nameRemedy Requester Console/app-name
  /prefetch-app
  prefetch-app
app-nameRemedy Task Management/app-name
  /prefetch-app
  prefetch-app
app-nameRemedy Reporting Console/app-name
  /prefetch-app
  /prefetch-server
  prefetch-server
server-nameremedy6.ars.unt.edu/server-name
prefetch-app
app-nameRemedy Foundation Elements/app-name
  prefetch-form
 form-nameSHR:OverviewConsole/form-name
  /prefetch-form
  /prefetch-app
  /prefetch-server
  prefetch-server
server-nameremedy6.ars.unt.edu/server-name
  prefetch-form
 form-nameHome Page/form-name
  /prefetch-form
  prefetch-form
 form-nameSHR:OverviewConsole/form-name
  /prefetch-form
  prefetch-form
 form-nameHPD:Help Desk/form-name
  /prefetch-form
  prefetch-form
 form-nameCHG:Infrastructure Change/form-name
  /prefetch-form
  prefetch-form
 form-namePBM:Known Error/form-name
  /prefetch-form
  prefetch-form
 form-namePBM:Problem Investigation/form-name
  /prefetch-form
  prefetch-form
 form-namePBM:Solution Database/form-name
  /prefetch-form
  prefetch-form
 form-nameTMS:Task/form-name
  /prefetch-form
  /prefetch-server
/prefetch-user
  /midtier-prefetch-config
 
  I am not convinced that everything in the script is in fact
 prefetching
  yet. It takes over 30 minutes for prefetch to run after the mid-tier
  server starts (the mid-tier shows little CPU usage but the AR Server
  sits at 30% utilization throughout!), but the difference in
 performance
  for the first time a form loads can be huge. Also, that difference
  varies based on who is logging in - the script was run with Demo,
 tests
  were with a support staff login (AREA authenticated) and then with the
  appadmin account, which was slower:
 
  Console IncidentChange Problem   Overview
  Requester
 
  Not Prefetched load time  0:39   1:35   0:25   0:47
  0:26
  Prefetched, 1st time load 0:04   0:22   0:04   0:27
  0:08
  (same, but different user)0:16   0:53   0:12   0:08
  0:29
  All subsequent loadings 

Re: Mid-tier prefetch scripts

2007-01-18 Thread Kyle Whitley
The perl script never really worked well for me, and like Axton said I 
had to use a no-permission user to get my requester/self service views 
to cache correctly.  Guess I need to upgrade my midtier.


Kyle

Axton wrote:

**
I believe the prefetch.xml is driven by the permissions of the account 
used to perform the prefetch.  That's to say that using an admin 
account to perform the prefetch will not have as much of a benefit for 
non-permission users accessing the system.  I configured our servers 
to use a no-permission account to perform the prefetch, since that is 
the permission base of our end users that access the mid-tier interface.
 
Axton Grams


 
On 1/18/07, *Kyle Whitley* [EMAIL PROTECTED] 
mailto:[EMAIL PROTECTED] wrote:


Is this only for ARS/Mid Tier 7.x  and ITSM 7.x ?


The only prefetch script I have seen is a perl script  not
anything in
xml.  I will say that with the perl script and my testing(ITSM 6
and ARS
6.3), load times  really depended on privileges of the user that was
used to authenticate to the AR System in the script.  Meaning if I
used
an Admin login in the script the admin views loaded quicker than a
view
a support person would receive on initial load of the view.  I
experienced similar load times using the perl script in ARS 6.3,
but my
runtime of the perl script was shorter,  I wasn't loading as many
forms
as you have.  I only ran the script on requester views for self
service.

Kyle

strauss wrote:
 Does anyone think that they have figured out all of the ins and
outs of
 using the prefetch script with mid-tier ( prefetchConfig.xml)??
Since
 this is a largely undocumented utility that BMC Remedy Support
does not
 fully understand, I'll take any hints that others have already
figured
 out. The latest script I am trying to use for ITSM 7.0 looks
like this:

 ?xml version=1.0 encoding=UTF-8?

midtier-prefetch-config  xmlns=http://www.bmc.com/remedy/midtier/700
http://www.bmc.com/remedy/midtier/700
   prefetch-user
 user-nameDemo/user-name
 localeen_US/locale
 prefetch-server
   server-name remedy6.ars.unt.edu
http://remedy6.ars.unt.edu/server-name
 prefetch-app
   app-nameRemedy Incident Management/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Change Management/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Change Management Dashboard/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Problem Management/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Requester Console/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Task Management/app-name
 /prefetch-app
 prefetch-app
   app-nameRemedy Reporting Console/app-name
 /prefetch-app
 /prefetch-server
 prefetch-server
   server-nameremedy6.ars.unt.edu
http://remedy6.ars.unt.edu/server-name
   prefetch-app
   app-nameRemedy Foundation Elements/app-name
 prefetch-form
form-nameSHR:OverviewConsole/form-name
 /prefetch-form
 /prefetch-app
 /prefetch-server
 prefetch-server
   server-nameremedy6.ars.unt.edu
http://remedy6.ars.unt.edu/server-name
 prefetch-form
form-nameHome Page/form-name
 /prefetch-form
 prefetch-form
form-nameSHR:OverviewConsole/form-name
 /prefetch-form
 prefetch-form
form-nameHPD:Help Desk/form-name
 /prefetch-form
 prefetch-form
form-nameCHG:Infrastructure Change/form-name
 /prefetch-form
 prefetch-form
form-namePBM:Known Error/form-name
 /prefetch-form
 prefetch-form
form-namePBM:Problem Investigation/form-name
 /prefetch-form
 prefetch-form
form-namePBM:Solution Database/form-name
 /prefetch-form
 prefetch-form
form-nameTMS:Task/form-name
 /prefetch-form
 /prefetch-server
   /prefetch-user
 /midtier-prefetch-config

 I am not convinced that everything in the script is in fact
prefetching
 yet. It takes over 30 minutes for prefetch to run after the
mid-tier
 server starts (the mid-tier shows little CPU usage but the AR Server
 sits at 30% utilization throughout!), but the difference in
performance
 for the first time a form loads can be huge. Also, that 

Re: Mid-tier prefetch scripts

2007-01-18 Thread Chris Akens
From what I understand, the mid-tier pre-caching is actually done on a
unique permission group list basis. Therefore you would want to pre-cache
with an account that has the same permissions as the majority of your users.
I have not used the pre-cache functionality that is included as part of 7.0;
but am interested in how it takes into account the different permission
group combinations. Any insight into this matter would be appreciated.

I have been less than impressed with the mid-tier v7 thus far to say the
least. Any suggestions on tuning the mid-tier for performance?

___
UNSUBSCRIBE or access ARSlist Archives at www.arslist.org ARSlist:Where the 
Answers Are