OPS/MVS

Hi,

I have question

When i have setup PAGSLOT% in threshold it will counts value  according red marked field ? 

Because i am wondering if i have threshodl setup to 40% alert is triggered. but when  i have every single datasets below 35% (blue marked) shouldn't it be less then 40% ? am I missing something ? Can anyone tell me how it's works ? 

-Br M 

Hello,

As is stated in the HELP for the command:

LOCAL page data set totals.
Total Total LOCAL size in bytes.
InUse Used LOCAL size in bytes followed by the percent of the total that is in use.  

The percentages are mathmetically derrived from the Used size of the page in bytes / Total Size of the page in bytes. 

I.E. 

8.33G Used divided by 20.5 G Size equals approximately 40.61%

The same holds true for the individual Page DS names.

I.E. PLPA

78.4M Used divided by 468M Size equals approximately 16.73 % used.

Hi,

Info what have you send is clear  but i still missing some explanation  in the datail list of every LOCAL type 

So i will try it one more time and better:

there is five datasets -all together have size 20,5G and EVERY single one have ~35% Pct% usage and SUM of Used should be 6,95 

but in TOTAL are wrong data 

If there was one more page it would match but know there is bug . see below Can you explain this ?

Another example from another lpar:

Both Lpar have same sysview package level 

-Br M 

Additionally, 

The LOCAL Used that is shown in the Info Area is the sum of the NONVIO and VIO slots used, both obtained directly from the ASMVT (Aux Storage Manager Vector Table) data area.

The Used slots for each LOCAL page dataset is obtained from Page Activity Reference Table (PART) data area. 

Hello Martin,

Investigating this further there is currently an open case with the same problem that you are reporting.

However, once our Sustaining Engineering team looked into this further the customer was recommended to open a PMR with IBM to investigate the discrepency. 

The suggestion was the problem description should state that there is a discrepancy between the total LOCAL slot usage in the ASMVT (ASMVSC + ASMNVSC) and the sum total in-use slots for each LOCAL page dataset (PARESZSL – PARESLTA).  

IBM is currently investigating and they are seeing oddities as well but have nothing conclusive to date.

-Bob

Hi,

cau you give me some link to opened case. to see what is the result or solution to this issue ? 

Hello Martin,

For the other open case, the customer opened a PMR with IBM (93060,130,702) . IBM was unable to determine the problem and after an IPL, the local slot count discrepancy disappeared. I would suggest opening a PMR and reference the other customer's case.