Conference ssdevo::hsz40_product

>    Is SWEAT available for V3.0Z?

    Yup...I'm  still  adding  bits  to  it  (Host  based  Raid  and  new 
    controller support), but the latest saveset is available at:-

    MAJERE::DISK$PUBLIC:[PUBLIC.STORAGEWORKS.TOOLS]SWEAT030.A

    ...this will be updated as and when I get time.

    
    Hello all,
    
    Sweat030 is ok with HSOF V3.0Z,you see the device locator failed in a
    stripe,mirror or raid set.
    
    The only error that I saw is for the version.Sweat see Cronic V8.00.
    
    Best regards.
    
    Olivier Cavailhe.

BTW, what's SWEAT?

    
    	(Chris, how about some fan-mail..... ;-)
    
    S W E A T = Storage Works Errorlog Analyzing Tool. Created by Chris
    Loane. It is MUCH better in translating Storageworks (and other SCSI
    related, or better tape/disk related) errorlog entries than ANA/ERR
    or DECevent. I use it all the time. Go get it, you'll love it. No
    sweat.
    
    	Bart Rietkerk.
    

>BTW, what's SWEAT?

    It's  the  StorageWorks  Errorlog  Analyser  Tool; a tool written to 
    bust out Binary Errorlog details from O'VMS systems.

Any version for UNIX?

    'fraid   not;   I   haven't   figured   out  the  Unix  errorlogging 
    peculiarities yet. Besides which, nowadays DECevent is probably  the 
    best bet.

Unfortunately the decevent folks don't seem to have plans for this
anytime soon...what info do you need about errorlog entries on unix?

If all it does is "bust out Binary Errorlog details", then
DECevent already does it.  So unless it really does "analysis"
then it's probably just another bit to text translator with
some new format of output.

For example, DECevent already tells you the device locator for
any HSZ40/50/... stripe/mirror/raid/jbod related errors.

Part of the problem on VMS is that it didn't report that stuff
completely.  DECevent on UNIX did, so this tool may be a no-op
for UNIX systems.

    Fred,
    
    You are correct. SWEAT provides for translation of Array Controller
    error log entries to a readable text format. It was required for
    support of HSD and HSJ controllers on VAX systems running VMS. Since
    DECevent performs the BTT (bit to text) translation on Alpha systems,
    it was only a VAX tool. With DECevent on D'UNIX, there is no real
    need for SWEAT in that enviroment. 
    
    I am not aware of any analysis ability of SWEAT, I do not think that
    analysis will be included. If analysis is required by the field
    service engineer, they should request that from the DECevent tools
    group.
    
    Bert
    

>    I am not aware of any analysis ability of SWEAT

    SWEAT  is purely BTT; however it does have 3 useful features (from a 
    Field engineer's point of view):-
    - The ability to merge and chronologically order  multiple  errorlog 
    files
    -  Single  line per event output (as in Stu Fuller's FSTERR tool and 
    TOPS20's SYSERR/BRIEF!!).
    - The ability to feed it a CONFIG file so that  SWEAT  reports  real 
    controller names (instead of something meaningless like 0009DEADBEEF)

>    If analysis is required by the field service engineer, they should 
>    request that from the DECevent tools group.

    Agreed!!

    > It was required for
    > support of HSD and HSJ controllers on VAX systems running VMS. Since
    > DECevent performs the BTT (bit to text) translation on Alpha systems,
    > it was only a VAX tool.
    
    Does it run on OpenVMS Alphs also???
    
    Dave

>    Does it run on OpenVMS Alphs also???

    It  runs  on  VAXen  and  Alphas;  it  can  merge ELE's from a mixed 
    cluster and then bust them out (you do NOT  have  to  run  an  Alpha 
    image to bust Alpha ELE's).