Highlighted
Absent Member.
Absent Member.
308 views

Oracle 2 Tier protocol script results are accurate?

Recently working on Oracle 2 tier I noticed that numbers from controller were
not jiving with the response time we experience when working on the application
via it's front end. We are working on an application that uses a fat client
(built in DotNet) and talks to Oracle 10g server using OCI (aka Oracle 2 tier)
protocol.

Based on that I have come to conclusion that the application spends a lot of
time at client and that does not get reflected in numbers rom conttoller. I
would like to hear from folks involved in testing any kind of two tier
database-client application with thick client.

Some more details are provided below.Now the thing I noticed about this
application is that a lot of records are being fetched. It could be something
like 3000-7000 record fetch per transaction. A transaction comprises of over
about 500 database invocations.

Now I can't understand why so many records would be required. However, what is
more important is what thick client is doing with them. If it is scrolling
through all the records of the result set and doing some parsing then I see it
causing significant client side latency contribution to overall response times.
The client contribution to the latency is something we don't see in numbers from
controller.
regards,
Vikram
0 Likes
3 Replies
Highlighted
Absent Member.. Absent Member..
Absent Member..

Re: Oracle 2 Tier protocol script results are accurate?

You are correct, Vikram. There would be rendering time on the client that would not be reflected in the LR results, because LR only has knowledge of timing between client and server. That is a current limitation for LR, not only in dealing with thick clients, but for all Web 2.0 applications. It is also something that HP is keenly aware of, and R&D is hard at work finding a way of closing that gap. Not sure if it will help you in the short run, though.
0 Likes
Highlighted
Absent Member.
Absent Member.

Re: Oracle 2 Tier protocol script results are accurate?

HI Vikram,

It will be great, if you light uo on this issye for me as i am also facing the same problem of response time mis match as of actual user and the lr transaction. When i am running the script as LR script response time is greater than the actual user while its less than the actual user experience when i am running it through BAC/BPM.

Please suggest as am not sure about the script results.
Thanks.

--Hanumant
0 Likes
Highlighted
Absent Member.. Absent Member..
Absent Member..

Re: Oracle 2 Tier protocol script results are accurate?

@ Hanuman,

depending on logging level, transaction could be longer, because time is count inside the transaction.
But depending on the script itself, some action can take time and is also count inside the transaction.
I open a thread about this problem (I think) on the web_reg_find function but some other could have such behaviour:
http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1371948&admit=109447626+1253181396192+28353475

@Vikram,

I experience the same trouble with oracle 2-tiers protocol, PC have reveal approxmitivelly 16000 event, which means about 3000 request to the database just for a logging !


--
david
0 Likes
The opinions expressed above are the personal opinions of the authors, not of Micro Focus. By using this site, you accept the Terms of Use and Rules of Participation. Certain versions of content ("Material") accessible here may contain branding from Hewlett-Packard Company (now HP Inc.) and Hewlett Packard Enterprise Company. As of September 1, 2017, the Material is now offered by Micro Focus, a separately owned and operated company. Any reference to the HP and Hewlett Packard Enterprise/HPE marks is historical in nature, and the HP and Hewlett Packard Enterprise/HPE marks are the property of their respective owners.