Using a file as alternate index without defining it as AIX

siddhuhandsome · New User Joined: 28 Feb 2004 Posts: 12 Location: Hyderabad

Hi everyone,

I have a doubt.
In our application, we have two datasets SERVICEADDRESS
and SERVICEALTX.
Physically the SERVICEALTX dataset is not defined as an AIX but logically am using it as an alternate index to the SERVICEADDRESS dataset.
SERVICEALTX is defined as a vsam ksds cluster.
Can someone explain how this is possible. Without defining an AIX, without relating it , without building a path. How is this possible?

The key in the SERVICEALTX is account number which is a part of SERVICEADDRESS. But both these are not related. Both are separate clusters.

bluebird · Active User Joined: 03 Feb 2004 Posts: 127

1 did u do a listc that would confirm your saying.
2 I believe this may be feasible thru programming e.g have the two datasets used in the app.

siddhuhandsome · New User Joined: 28 Feb 2004 Posts: 12 Location: Hyderabad

Snapshot of ListC for SERVINDX
ASSOCIATIONS
DATA-----$3S000.F001.SERVINDX.DATA
INDEX----$3S000.F001.SERVINDX.INDEX
DATA ------- $3S000.F001.SERVINDX.DATA
IN-CAT --- UCAT.CISCAT

Snapshot of Listc for SERVALTX

ASSOCIATIONS
DATA-----$3S000.F001.SERVALTX.DATA
INDEX----$3S000.F001.SERVALTX.INDEX
DATA ------- $3S000.F001.SERVALTX.DATA
IN-CAT --- UCAT.CISCAT

In the JCL where these two are defined, I found that they are defined as clusters. Also it is evident from the listc as above, that both are independent datasets. I don't understand how they both are being used in such a manner.

bluebird · Active User Joined: 03 Feb 2004 Posts: 127

hello,

sorry for the delay...

u should do a listc ent(/) all on the path dataset (it is the link between cluster and alternate index yes that's what a path is 4)

siddhuhandsome · New User Joined: 28 Feb 2004 Posts: 12 Location: Hyderabad

Hi,

I have done the LISTC ENT(/) ALL. But to show it in short I took the snapshot of the datasets dependents.

Also there is no path for the dataset. So I cannot try the listc ent(/) all for the path dataset.

mmwife · Super Moderator Joined: 30 May 2003 Posts: 1592

Hi Sid,

That technique was used quite often in the days before VSAM and ISAM using direct datasets.

It's still used today when performance of the VSAM ALT INDEX processing is an issue.

The pgmr assumes responsibility for the coding to implement the technique. We used to call it "RYO" (roll your own).

Regards, Jack.

siddhuhandsome · New User Joined: 28 Feb 2004 Posts: 12 Location: Hyderabad

Hey Jack,

Thanks for your response

Regards,
Siddhu