Tom,
I think you misunderstood, there is a parameter, MIGPR, on the storage
pool definition, that effectively limits the number of tape drives
migration will use. But, now that I think about it, that wait would not
count toward your mount wait count. So I will re-lurk and defer to Mr.
Colwell, who appears to have a likely solution.
Ron 'Is ADSM really that cool, or do I just need to get out more?'
Clendenny
Callaway Nuclear Plant
Fulton, Missouri
<rdclendenny AT cal.ue DOT com>
> -----Original Message-----
> From: Tom Tann{s [SMTP:tom.tannas AT USIT.UIO DOT NO]
> Sent: Thursday, October 30, 1997 9:13 AM
> To: ADSM-L AT VM.MARIST DOT EDU
> Subject: Re: migration/mount waits
>
> On Thu, 30 Oct 1997, Clendenny, Ronald D. wrote:
>
> > Tom,
> > I have seen some very erroneous wait times shortly after midnight,
> which
> > I attributed to a math/logic error in calculating the time
> difference
> > across 2 days. The wait times later in the day were correct.
> > Also, how many processes is your migration limited to. That will
> > determine how many drives you use in the 3494 and, therefore, cause
> a
> > wait, even if other drives are not busy.
> > BTW, we also have a 3494 w/ 4 3590's (the last 2 were installed last
> > night). Its a lot more fun with 4 drives!
> >
> > Ron Clendenny
> > Callaway Nuclear Plant
> > Fulton, Missouri
> > <rdclendenny AT cal.ue DOT com>
> >
> No, this is not a "midnight" problem. The example I posted was taken
> from
> a process that was running from 9 to 11.30 this morning.
>
> Yes, other activity in the library will have effect on the wait-time.
> But the real "wait time" is still around the average of 60-90 secs.
> +/-
> 50%.
|