[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference noted::decps

Title:	POLYCENTER Performance Data Collector & Advisor for OpenVMS
Notice:	Latest version V2.2-51
Moderator:	BSS::JILSON

Created:	Wed Nov 07 1990
Last Modified:	Thu Jun 05 1997
Last Successful Update:	Fri Jun 06 1997
Number of topics:	2413
Total number of notes:	9017

2387.0. "Channels with HSJ based mirrorsets" by KERNEL::TITCOMBER () Tue Feb 11 1997 05:00

    Hi,

    I have a customer who is running DECPS v2.1-48 on an OpenVMS/VAX V6.1
    cluster.

    The cluster consists of three VAX 7810 systems and six HSJ40s, which
    are configured in a dual redundancy mode and running HSOF V2.7.  The
    drives on the HSJ40's are a mixture of RZ28-VA, RZ28D-VA, RZ29B-VA,
    RZ29B-VW and RZ28D-VW.  These drives are spread across 12 BA350 shelves
    and 6 BA356 shelves with 8-bit personality modules.  The VW (sixteen
    bit variants) are only on the BA356 shelves.

    They have recently introduced HSJ based disk mirrorsets and this has
    produced some unusual results within DECPS.  The mirrorsets are spread
    across a number of shelves (channels?) and these are configured as
    follows - Channel 0 & 3, 1 & 4, 2 & 5.

    The customer believes that if DECPS was properly reporting on the
    mirrorsets then it would display channels such as HSJ014_03, _14 & _25.

    However, graphs and reports for channel utilisation are now showing
    channels HSJ004_15, HSJ005_15 (etc).

    Can anybody enlighten me about how DECPS constructs these channel
    numbers for HSJs? (i.e. where does the '_15' comes from?)

    Thanks in advance,

    Rich

T.R Title User Personal
Name Date Lines

2387.1 BSS::JILSON WFH in the Chemung River Valley Tue Feb 11 1997 09:14 8

T.R	Title	User	Personal Name	Date	Lines
2387.1		BSS::JILSON	WFH in the Chemung River Valley	`Tue Feb 11 1997 09:14`	8
	PSDC collects just what VMS tells it and VMS doesn't care to make sure that what it is reporting is 100% correct as it is not information needed for any IO function. IMHO VMS engineering will not expend any resources to make sure these 'cosmetic' items are correct (same with Alternate host name in SHOW DEV/FULL ddcu:). Your only option is to use the PSDC logicals for disk to HSC mapping and see if they effect these HSJ based sets. Jilly

PSDC collects just what VMS tells it and VMS doesn't care to make sure that 
what it is reporting is 100% correct as it is not information needed for 
any IO function.  IMHO VMS engineering will not expend any resources to 
make sure these 'cosmetic' items are correct (same with Alternate host name 
in SHOW DEV/FULL ddcu:).  Your only option is to use the PSDC logicals for 
disk to HSC mapping and see if they effect these HSJ based sets.

Jilly