ADADBS utility

Device: 8390
Trks/Cyl: 15
ASSO: 3440:14
DATA: 6518:8
Work: 10706:5
PLOG/RLOG: 10706:5
CLOG: 8904:6
TEMP/SORT/DSIM:8904:6
Notes: 1

I need to setup a job to increase the number of ISN by 12000 for DB, FILE xx. ISN is stored in AC.
I need to use ADADBS utility to find out which utility function to use.

I think it is

ADADBS ALLOCATE: Allocate File Extent

Is it correct?

Any help will be appreciated.

thanks

Thanks

Yes, this is the correct utility function to allocate a additional AC extent,
but your terminology (“ISN is stored in AC”) is incorrect, the AC does not
hold ISNs but the ISN is the index into the AC to find the associated DATA block.

your terminology (“ISN is stored in AC”) is incorrect, the AC does not
hold ISNs but the ISN is the index into the AC to find the associated DATA block.

can you please refer some documents then I can try to convince my lead on it?

thx

what about ADADBS INCREASE DATASIZE/ASSOSIZE?

what is diff btw ADADBS ALLOCATE & ADADBS INCREASE

If it is ADADBS ALLOCATE then i sit ok?

ADADBS ALLOCATE FILE= 55
{ACSIZE | AC2SIZE | DSSIZE| NISIZE | UISIZE} = size
[DEVICE= 8390]
[NOUSERNAND]
[PASSWORD]
[STARTRABN= star-rabn] ??
[TEST]

from where I will get values for {ACSIZE | AC2SIZE | DSSIZE| NISIZE | UISIZE?

Is from report?

what is ACSIZE , is it the size of AC?

thx

Formula in the documentation here

And you can extrapolate from the current sizes (from adarep) re. what you will
need for additional “n” records.

ALLOCATE provides additional space for the components of an existing file within the
allocated cotainers (ASSO / DATA)

INCREASE adds space to the container (ASSO / DATA) itself.

If RABNSIZE=3, MAXISN=5000 on a model 3380 with 668 entries per block (2004/3) results in 8 blocks. The highest ISN expected (before further expansion) is therefore 5343 (668 * 8 - 1).

If RABNSIZE=4, MAXISN=5000 on a model 3380 with 501 entries per block (2004/4) results in 10 blocks. The highest ISN expected
is therefore 5009 (501 * 10 - 1).

http://techcommunity.softwareag.com/ecosystem/documentation/adabas/ada824mfr/adamf/dbacard/dballoc.htm

From where are they getting 5343 and 5009 values in 1 & 2 respectively.
Why did they deduct -1 in both 1 & 2?

In my case

Device----Trks/Cyl---- ASSO------ DATA---------- WORK--------- PLOG/RLOG--------- CLOG------- TEMP/SORT/DSIM-------- Notes
8390----- 15 --------3440:14----- 6518:8-------- 10706:5--------- 10706:5----------- 8904:6----------- 8904:6------------------- 1

By Using the information provided to calculate how many RABN will be required to increase number of ISN by 12000. Each ISN takes 3 byte.

3*12000/3440 = 10.465 ,
is it a block number?

to do further what I need it?

thanks
file55.pdf (88.4 KB)

I assume the -1 is due to a pointer linking the extents.

In your case:

3440/3 = 1146
12000/1146=10,47

so you will need 11 AC RABNs for your 12000 ISNs, in fact your MAXISN will go up by 12605.

But as I said, the AC is not the only component to consider.

‘-1’ because the first entry in address converter represents ISN = 0 (which does not exists).

I think next step I need to write JCL as

ADADBS ALLOCATE FILE= 55
{ACSIZE | AC2SIZE | DSSIZE| NISIZE | UISIZE} = size
[DEVICE= 8390]
[NOUSERNAND]
[PASSWORD]
[STARTRABN= star-rabn]
[TEST]

One of the example I read in the manual as

Example
An address converter extent of 30 blocks is to be allocated for file 15.

ADADBS ALLOCATE FILE=15,ACSIZE=30B

  1. Do I need to put one value for all as:
    {ACSIZE | AC2SIZE | DSSIZE| NISIZE | UISIZE} =30B

Is 30B same as block length? where should I look for extent of block in the ADAREP?
Is it from .P H Y S I C A L L A Y O U T ?

Extents in Blk
From To

1 314986

{ACSIZE | AC2SIZE | DSSIZE| NISIZE | UISIZE} =314986

30B - ‘B’ stands for ‘blocks’ and the number is the amount you want to extent - this is documented in manual.

You can also let nucleus (automatic) allocate additional extents.

Quote from the documentation:

what about [STARTRABN= star-rabn] ?

ListType----From ------ To
RABN-------RABN
AC ----------71023-----71284

Do i need to use 7123 or 11?

Forget about STARTRABN, one usually does’t care about where an extent is created,
only in very, very rare and special cases where one needs to fine-tune access path
with multiple volumes connected over multiple channels, etc. etc. …

But with todays disk architectures I don’t think there’s still a reason to even look into
doing this.

what about this calculation?
Where should I use?

3440/3 = 1146
12000/1146=10,47

so you will need 11 AC RABNs for your 12000 ISNs, in fact your MAXISN will go up by 12605.

Mu goal is to Setup a job to increase the number of ISN by 12000. ISN is stored in AC.

thanks

Then use ACSIZE=11B as per that calculation.

ADADBS ALLOCATE FILE= 55
{ACSIZE | AC2SIZE | DSSIZE| NISIZE | UISIZE} = 11B
[DEVICE= 8390]
[NOUSERNAND]
[PASSWORD]
[STARTRABN= 71023] – I got this value From RABN for list type AC

As I said before, this just is not true.

The AC only maps the ISN to the physical data block for that ISN.

From the documentation (a bit into that section of the documentation):

No, please have a look at the conventions section re. what square brackets, curly braces etc. denote.