Veritas-bu

[Veritas-bu] Drive Utilization

2002-03-28 17:12:09
Subject: [Veritas-bu] Drive Utilization
From: DDobbs AT mutualmaterials DOT com (Dan Dobbs)
Date: Thu, 28 Mar 2002 14:12:09 -0800
Is there a way to compute 'real' backup times, however? If I have 8 streams 
that all start at the same time (different mount points), but four drives, I 
get something like:

#./bpdbjobs -report | more

     13291   Backup     Done      0      MMORA1-Oracle    MMORA1-O.Full         
mmora1       MMORA3   03/19/2002 21:17:15    01:14:27      03/19/2002 22:31:42  
     3725683
     13290   Backup     Done      0      MMORA1-Oracle    MMORA1-O.Full         
mmora1       MMORA3   03/19/2002 21:17:15    02:01:47      03/19/2002 23:19:02  
    17798378
     13289   Backup     Done      0      MMORA1-Oracle    MMORA1-O.Full         
mmora1       MMORA3   03/19/2002 21:17:15    00:11:44      03/19/2002 21:28:59  
      300266
     13288   Backup     Done      0      MMORA1-Oracle    MMORA1-O.Full         
mmora1       MMORA3   03/19/2002 21:17:15    02:13:39      03/19/2002 23:30:54  
     8376030
     13287   Backup     Done      0      MMORA1-Oracle    MMORA1-O.Full         
mmora1       MMORA3   03/19/2002 21:17:15    00:34:31      03/19/2002 21:51:46  
     4422003
     13286   Backup     Done      0      MMORA1-Oracle    MMORA1-O.Full         
mmora1       MMORA3   03/19/2002 21:17:15    00:11:27      03/19/2002 21:28:42  
      245566
     13285   Backup     Done      0      MMORA1-Oracle    MMORA1-O.Full         
mmora1       MMORA3   03/19/2002 21:17:15    02:17:51      03/19/2002 23:35:06  
     7742698
     13284   Backup     Done      0      MMORA1-Oracle    MMORA1-O.Full         
mmora1       MMORA3   03/19/2002 21:17:15    00:04:19      03/19/2002 21:21:34  
      182804

Obviously some didn't start really doing anything until a tape drive became 
available--can one figure out the 'true' backup time?

-dd

>>> "Donaldson, Mark" <Mark.Donaldson AT experianems DOT com> 03/28/02 01:16PM 
>>> >>>
The "bpdbjobs -report" has all of this except which drive was utilized.  

Also checkout "bpdbjobs -report -all_columns".  

See the FAQ in Storage Mountain for a description of the columns.

http://www.backupcentral.com/cgi-bin/netbackup-fom?_highlightWords=bpdbjobs&; 
file=194

Anyway, bpdbjobs has start time, stop times, KB transferred (calc to get
rates), server, class, etc.

To get the drive information, you have to go parse the database.

-Mark

-----Original Message-----
From: Jeff A. Earickson [mailto:jaearick AT colby DOT edu] 
Sent: Thursday, March 28, 2002 1:40 PM
To: Gary Sperano
Cc: veritas-bu AT mailman.eng.auburn DOT edu 
Subject: Re: [Veritas-bu] Drive Utilization


Hi,
   Funny, but I needed the same info the other day, when I wanted to know
what machine/class was backing up to what drive, when, and what tape it
was writing to.  Hey Veritas, how come no nice report to tell you?

* start time of backup job
* stop time of backup job
* elapsed time and thruput rate
* what drive got used and what tape got written to

Really cool would be a GUI barchart that showed, on a per drive basis,
machine backups over time.

Anyway, I cooked up the attached shell script to go thru every class,
show what happened in the previous 24 hours (weeding out trivial messages),
and give me real timestamps as well.  The output is not pretty but the
info is all there for a backup or restore.  Output looks like:

============== diamond
Wed Mar 27 18:07:35 2002 added backup job for client diamond, class diamond,
schedule Full part 0 to NetBackup scheduler work queue
Wed Mar 27 18:25:25 2002 started backup job for client diamond, class
diamond, schedule Full on storage unit slate-dlt-robot-tld-0
Wed Mar 27 18:25:27 2002 client diamond handling path ALL_LOCAL_DRIVES
Wed Mar 27 18:25:39 2002 begin writing backup id diamond_1017271525, copy 1,
fragment 1, to media id QWC775 on drive index 1
Wed Mar 27 19:10:09 2002 successfully wrote backup id diamond_1017271525,
copy 1, fragment 1, 8677536 Kbytes at 3271.484 Kbytes/sec
Wed Mar 27 19:10:11 2002 CLIENT diamond CLASS diamond SCHED Full EXIT STATUS
0 (the requested operation was successfully completed)
Thu Mar 28 10:18:54 2002 Media id QZF816 is needed for the restore.
Thu Mar 28 10:24:28 2002 begin reading backup id onyx_1015153648 (restore),
copy 1, fragment 1 from media id QZF816 on drive index 1
Thu Mar 28 10:24:29 2002 successfully restored from backup id
onyx_1015153648, copy 1, 160 kbytes

** Jeff A. Earickson, Ph.D                         PHONE: 207-872-3659
** Senior UNIX Sysadmin, Information Technology    EMAIL: jaearick AT colby DOT 
edu 
** Colby College, 4214 Mayflower Hill,               FAX: 207-872-3076
** Waterville ME, 04901-8842
----------------------------------------------------------------------------

On Thu, 28 Mar 2002, Gary Sperano wrote:

> Date: Thu, 28 Mar 2002 14:43:14 -0500
> From: Gary Sperano <gsperano AT powertel DOT com>
> To: veritas-bu AT mailman.eng.auburn DOT edu 
> Subject: [Veritas-bu] Drive Utilization
>
>
>
> Does anybody have any good scripts (ksh, perl, etc) that reports drive(s)
> utilization or know where to obtain this type of process from?
>
> Gary A. Sperano Jr
> Technical Specialist
> VoiceStream
> PCS: 404-610-9566
> Desk: 770-604-3165
> Gsperano AT powertel DOT com 
>
>
> _______________________________________________
> Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu 
> http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu 
>

<Prev in Thread] Current Thread [Next in Thread>