Capacity and demand

[Sarb_Singh]

Hi team,

I have been asked if Prognosis can provide the following data:

Number of incoming comms stations
Number of APACS processes
Number of VISA processes
Number of Visa stations
Number of MasterCard stations
Number of MasterCard stations
Number and capacity of HSMs
Placement of data on discs so that peak IO rate not exceeded
Understanding of interaction of batch and online processing

Could you please inform me of what data fields I would need?

Thank you very much.

[ChristopherS]

Hi Sarb,
At a high level yes most likely to most of not all of these. .. there are some clarifying questions though to provide something more specific in terms of how exactly we would do it in your environment.

If XPNET, it is fairly straight forward for the stations and processes. If this is in ICE/XS or Base24eps / Linux we need to look a bit deeper... if on IBM System Z.. then we'd need to really get creative.

If you want details from within the APACS or IS processes (not just a count or if they are up/down) like to discuss as we know how to do it but there is nothing currently out of the box but are building up a list of customers interested in this information.

HSMs.. it depends on which brand of HSM you have on the level of effort as like for example HP Atalla's Boxcar provides very little information unless in trace mode but some other platforms provide both a proxy with stats or the ability to potentially get stats right off of the HSM itself. We've looked at this for a couple of customers for a couple of different HSMs but it has been a while.

I know you don't run Postilion but we provide most of these by default in Postilion which is Windows based; except batch processing currently but it is something we are working on to monitor through its central batch monitoring process.

Batch processing on any platform requires a bit more discussion because there are a lot of different batch processes that run and log to different places from BIN / Routing Loading to Card / Account Refreshes and Extracts/Recons.

Please reach out to me or your account representative and they will know how to get a hold of me to dive into this in detail.

[Sarb_Singh]

Hi Christopher,

thank you for the reply. I am going to have to do some serious research

Christopher R Souser said
Hi Sarb,
At a high level yes most likely to most of not all of these. .. there are some clarifying questions though to provide something more specific in terms of how exactly we would do it in your environment.

If XPNET, it is fairly straight forward for the stations and processes. If this is in ICE/XS or Base24eps / Linux we need to look a bit deeper... if on IBM System Z.. then we'd need to really get creative.

XPNET

If you want details from within the APACS or IS processes (not just a count or if they are up/down) like to discuss as we know how to do it but there is nothing currently out of the box but are building up a list of customers interested in this information.

Just up or down and how many

HSMs.. it depends on which brand of HSM you have on the level of effort as like for example HP Atalla's Boxcar provides very little information unless in trace mode but some other platforms provide both a proxy with stats or the ability to potentially get stats right off of the HSM itself. We've looked at this for a couple of customers for a couple of different HSMs but it has been a while.

The HSM's are HP 9000's (will check further)

I know you don't run Postilion but we provide most of these by default in Postilion which is Windows based; except batch processing currently but it is something we are working on to monitor through its central batch monitoring process.

Batch processing on any platform requires a bit more discussion because there are a lot of different batch processes that run and log to different places from BIN / Routing Loading to Card / Account Refreshes and Extracts/Recons.

Please reach out to me or your account representative and they will know how to get a hold of me to dive into this in detail.  

I discussed with IR support who stated that this may cause significant overhead. So I would probably run a report once a week/month.

[ChristopherS]

Sarb Singh said
I discussed with IR support who stated that this may cause significant overhead. So I would probably run a report once a week/month.  

Actually most of it should absolutely NOT cause much of any overheard if we leverage information from the payment application or do it in an efficient way and do similar things on a couple of platforms this way. What would be extremely inefficient and cause some overhead is we had to resort to using something like a NETSTAT to get the station connection information but most if not all of it should be able to be collected and actually populated into a realtime dashboard and be able to threshold on it.. though not saying its necessarily out of the box (but at least some or potentially a majority is likely there in some form).

Christopher