1753602 Members
6983 Online
108796 Solutions
New Discussion юеВ

Re: History ECP

 
SOLVED
Go to solution
Wim Van den Wyngaert
Honored Contributor

History ECP

1) Does anyone knows the history of VPA / ECP ?
What is the difference ? Is ECP a superset of VPA ?
2) I have VPA 2.4 and ECP is at 5.4D. Are there really that many versions between them ?
What do I gain with an upgrade to ECP ?

Wim
Wim
5 REPLIES 5
John Eerenberg
Valued Contributor
Solution

Re: History ECP

A long time ago, at a company far, far away before ECP was even a TLA . . .
Digital created SPM -- a solid software performance monitor that gave accurate, straight forward performance stats without capacity planning functionality.
Then came VPA with its AI analyzer and a capacity planner, created by a group within Digital competing with SPM.
The two stored their respective performance metrics in incompatible formats; at least in the begginning.
The VPA capacity planner was rather lacking for those of us who wanted more from a CP product; it didn't collect several important details.
At the same time, DECcp (DEC Capacity Planner) was available and used the SPM data plus its own performance monitor/data collector (meant to be run for an hour or two for capacity planning purposes, though rumors of abuse exist) to capture key stroke information, etc. (i.e., information that neither SPM nor VPA would capture). This combination created the basis for a very accurate queueing network model which could then be sliced, diced, chopped, grated, grouped, categorized, upgraded and so any part of the computer could be modeled to determine what an upgrade (CPU, memory, disk controllers, and disks in any combination) would buy you.
So, one way to look at things was VPA leap-frogged SPM for performance stats purposes and provided one level of capacity planning, but DECcp leap frogged VPA's capacity planner. DECcp was dependent on SPM and, eventually, the needed SPM functionality was merged into VPA.
Then came the great Digital fire sale which disposed of VPA; CA bought VPA and called it TNG (actually VPA is but one component of TNG).
At this time SPM was more or less dead but appeared for a long time on the LP distribution.
So without Digital's support of the SPM portion of VPA, DECcp's development went into limbo land. CA controlled the SPM portion but didn't buy DECcp.
Some in Digital/Compaq/HP regretted the events of great fire sale (a whole 'nother story) and created ECP.
Now, ECP is a strange beast because the Compaq/HP version of it tried to become DECcp but also competed with Compaq/HP's AMDS (yet another whole 'nother story).
But, today, ECP is also a product from PerfCap Corp. which came from the highly regarded DECcp (i.e., the Digital developers of DECcp bought DECcp and started their own company and pushed the DECcp technology forward).

PerfCap's ECP is a superset of Compaq/HP ECP.

Okay. You for asked for the history, and as best I remember it, the above is pretty much on target and spans 15 if not 20 years.

> Is ECP a superset of VPA ?
Since there are two ECP's, and one TNG (i.e., VPA), the answer is "it depends."

> 2) I have VPA 2.4 and ECP is at 5.4D. Are there really that many versions between them ?
I would stick with just TNG/VPA and not worry about Compaq/HP ECP.

> What do I gain with an upgrade to ECP ?
My personal opinion is that the Compaq/HP ECP is not worth it; the PerfCap version of ECP (i.e., DECcp) is a complementary product to TNG/VPA (again, my opinion). A word of advice: don't use PerfCap's ECP by yourself unless you have a foundation in queueing theory; some details can be overlooked and can lead to incorrrect conclusions.

Summary:
I would use TNG/VPA for everyday performance measurements, graphs, trouble shooting performance, etc. In other words, system performance monitoring and tuning.
Once your system is free of any mistuning, artificial bottlenecks, etc. . . .
I would use PerfCap's ECP for planning "what if" senerios such as upgrading from EV6 ro EV7, replacing old HSZ controllers/disks with HSG or EVA technology, adding memory and how much, etc.

john
I do not receive any form of compensation from HP, CA, and/or PerfCap. The above is strictly my opinion; take what you like and leave the rest.
It is better to STQ then LDQ
Ian Miller.
Honored Contributor

Re: History ECP

sounds like a book could be written.

See also T4
http://h71000.www7.hp.com/openvms/products/t4/index.html
____________________
Purely Personal Opinion
Martin P.J. Zinser
Honored Contributor

Re: History ECP

Hello Wim,

usually the PerfCap folks are around at the bootcamp, the next of which will take place in May. Also the AMDS, T4, etc. folks are around. So if you need in-depth, first hand information including gossip on this, get a ticket to Nashuha (plus you can meet Hein and Sue ;-)

Greetings, Martin
Wim Van den Wyngaert
Honored Contributor

Re: History ECP

Yet another performance data collector ?

http://h71000.www7.hp.com/openvms/products/tdc/

What is the history of this one ?
Wim
Ian Miller.
Honored Contributor

Re: History ECP

I think tdc V1 was developed by VMS Eng to collect data to be processed by another companies product. I think tdc V2 is a move towards having one performance data collector for whatever needs to use it. TDC = The Data Collector.
____________________
Purely Personal Opinion