In the scenario I described below, collocation was turned off because of
the large amount of nodes the server supports.
The 86 tapemounts were the mounts from only the 26 tapes the data was
on.
ADSM writes everything to tape sequentially, and also looks for it in
that manner when it restores. If you have data on 5 tapes, it might
mount tape0, retrieve the info from that, then put in tape 2, retrieve
info from that, and then realize that it has more data on tape 0, mount
that again, and so on. I see it all of the time. As I said, UGH!
Jenn
> -----Original Message-----
> From: Hilton Tina [SMTP:HiltonT AT TCE DOT COM]
> Sent: Thursday, September 03, 1998 10:39 AM
> To: ADSM-L AT VM.MARIST DOT EDU
> Subject: Re: Performance problem
>
> Jennifer,
>
> Maybe I'm missing something obvious, but why would ADSM ask for a tape
> that
> it didn't need for the restore?
>
> Tina
>
> > -----Original Message-----
> > From: Cunningham, Jennifer
> [SMTP:Jennifer.Cunningham AT LIBERTYMUTUAL DOT COM]
> > Sent: Thursday, September 03, 1998 9:15 AM
> > To: ADSM-L AT VM.MARIST DOT EDU
> > Subject: Re: Performance problem
> >
> > I would also check to see how many tapes have data on them. You
> could
> > be looking at a lot of tape mounts...... Another dept here had a
> case
> > where it took 33 hours to restore 30G back to a UNIX box because of
> 82
> > tapemounts, but there was data on only 26 tapes. Ugh.
> >
> > Jennifer
> > ________________________
> > Jennifer Cunningham
> > Desktop Operations / Backup & Recovery
> > SDN #8-435-4029
> > 603-431-8400 x54029
> > Pager 1.800.759.4890
> > email pager - 7594890 AT skytel DOT com
> > jennifer.cunningham AT libertymutual DOT com
> >
> >
> >
> > > -----Original Message-----
> > > From: Paul Bergh [SMTP:pbergh AT ATT DOT NET]
> > > Sent: Thursday, September 03, 1998 10:50 AM
> > > To: ADSM-L AT VM.MARIST DOT EDU
> > > Subject: Re: Performance problem
> > >
> > > your problem is with the osa-2 adapter card, throughput is
> probably
> > > 2-3mb/sec, put
> > > additional osa cards in or acquire a ibm 2216 or equivalent.
> > >
> > > Tjeerd Saijoen wrote:
> > >
> > > Hello ADSM People I hope somebody can help me with this
> > > problem We have the following situation A IBM 9672/63 Mainframe
> with
> > > ADSM as server connected with a OSA-2 FDDI adapter to a Cabletron
> 9000
> > > smartswitch. >From the 9000 we have a 100MB Ethernet connected to
> a
> > > Novell server (NETWARE 3.12) we backing up 250.000 files. During
> the
> > > restore of the Novell server we restored 16 Gb of data in 25hrTape
> > > units are 3490E Here are the Netware
> > > parameters Nodename
> > > WBMAS COMMMETHOD
> > > TCPIPTCPSERVERADDRESS
> > > 128.10.1.20.201TCPPORT
> > > 1500TCPBUFFSIZE
> > > 32TCPWINDOWSIZE
> > > 64TXNBYTELIMIT 10240 DOMAIN
> ALL-LOCALDOMAIN
> > > DIRECTORYSCHEDLOG
> > > SYS:\ADSM\DSMSCHED.LOGSCHEDLOGRETENTION
> > > 7ERRORLOG
> > > SYS:\ADSM\DSMERROR.LOGERRORLOGRETENTION
> > > 4SCHEDMODE
> > > PROMPTEDTAPEPROMPT
> > > NO EXCLUDE
> .............. ------------
> > > OS/390 IUCV STARTLUname
> > > ........TCPName
> > > TCPIP2HTTPTCPPORT
> > > 1580COMMTimeout
> > > 600EXPINterval
> > > 0IDLETimeout
> > > 30MAXSessions
> 64TXNGroupmax
> > > 100BUFPoolsize
> > > 24576LOGPoolsize
> > > 1024VOLUMEHistory
> > > 'ADSM.SERV1.HIST1'DEVCONFig
> > > 'ADSM.DEVCONF1.DATA'DELetionexit
> > > ADSMTAPELICences CLIENTS 140LICences DESKTOPLICences UNIX Anyone a
> > > idea why it takes 25 hours to restore 16 Gb ?????? Regards Tjeerd
> > > SaijoenRiscmanagement tsaijoen AT riscman DOT com
> > > <mailto:tsaijoen AT riscman DOT com>
> > >
> > >
|
