Thanks, Itzik. Do you know if this problem exists in versions after
6.1.1? Unfortunately, while having only one available tape will work,
sooner or later, I'll have to load another tape, and it's much easier
for me to already have a couple extras loaded, so I guess it's really 6
of one half a dozen of another since it really doesn't matter what goes
to which since they're all part of the same clone pool. I mean, if you
have two tapes available then you'll fill them both half as fast as you
would if you had just one, but then if you had just one, it would last
half as long, so I guess "All roads lead to Rome" so to speak ... kind
of annoying, though.
Cheers.
George
Itzik Meirson wrote:
>
> George,
> Legato is still working on this issue so you will not find anything
> about this "in work" proposed patch/fix. You can open a call with Legato
> support, explain your problem/findings and then you may refer them to
> the mentioned fix.
> One workaround that I know is working is not to have the second tape
> available in the clone pool. In that case Networker will issue a
> mount/label request that cannot be satisfied. In the mean time the
> original tape will get idle and Networker will reuse it canceling the
> pending mount/label request.
> I hope this clarifies a bit the issue.
> Itzik
>
> > -----Original Message-----
> > From: George Sinclair [mailto:George.Sinclair AT noaa DOT gov]
> > Sent: Thursday, November 13, 2003 16:39
> > To: Legato NetWorker discussion; Itzik Meirson
> > Subject: Re: [Networker] Weirdness when cloning
> >
> > Thanks for the reply! Do you know where I can get more information on
> > this? What does "LGTpa49963" refer to? Is this a patch? Maybe
> something
> > Legato is still working on? I went to their web site and plugged that
> > into a search, but it turned up nothing. Do they discuss this
> somewhere
> > on their site?
> >
> > Sound like I'm just gonna have to tolerate NetWorker's decision on
> what
> > to write to which tape. Clearly, I have to have at least two available
> > volumes in case one fills up, and I'm not around to load another
> > requested tape. If NetWorker then decides to write one thing to one
> and
> > one thing to the other, rather than one tape at a time, so be it.
> >
> > Thanks.
> >
> > George
> >
> > Itzik Meirson wrote:
> > >
> > > Hi George,
> > > There are currently some mysterious issues in regard to tapes usage
> > > during cloning.
> > > I know it is handled currently under LGTpa49963 that is not
> > > closed/verified yet.
> > > This has to do with the time it takes the device/tape to become (be
> > > declared) idle after it completes the writing.
> > > Itzik
> > >
> > > > -----Original Message-----
> > > > From: George Sinclair [mailto:George.Sinclair AT noaa DOT gov]
> > > > Sent: Wednesday, November 12, 2003 21:39
> > > > To: NETWORKER AT LISTMAIL.TEMPLE DOT EDU
> > > > Subject: [Networker] Weirdness when cloning
> > > >
> > > > Hi,
> > > >
> > > > Seeing strange problem when cloning, not sure what to think! Not
> seen
> > > > this before.
> > > >
> > > > I have two available clone volumes, and when I run a clone
> operation,
> > > > NetWorker insists on cloning the saveset to one clone volume and
> the
> > > > index to the other clone volume rather than just sending
> everything to
> > > > one clone volume. It seems to do this when the ssids thatI'm
> cloning
> > > are
> > > > from different tapes.
> > > >
> > > > Here are the details: I have two clone pool volumes: ARC_c001,
> > > ARC_c002.
> > > > Both have plenty of space. I try to clone 4 savesets. 2 are on
> volume:
> > > > ARC001, and the other 2 are client indexes located on volume:
> ARC002.
> > > >
> > > > ARC001, and both clone tapes are on storage node library. ARC002
> is on
> > > > primary server's library.
> > > >
> > > > When I run clone command as: nsrclone -s server -b 'ARC Clone' -S
> > > ssid1
> > > > ssid2 ssid3 ssid4
> > > >
> > > > NetWorker clones two savesets to ARC_c001 (just what you'd
> expect),
> > > and
> > > > then when it starts to clone the two indexes it issues: "Device or
> > > > resource busy" message for the drive that has volume ARC0_c001. It
> > > then
> > > > loads ARC_c002 into another drive and clones the indexes to that
> tape,
> > > > not ARC_c001! If I then run the command again, using different
> ssids,
> > > > the same thing happens, but this time it writes the first two
> savesets
> > > > to ARC_c002 and then issues the "Device or resource busy" message
> on
> > > the
> > > > device containing ARC_c002 and then clones the indexes to
> ARC_c001. So
> > > > this is the reverse of the earlier. Subsequent tests just keep
> > > producing
> > > > the same swapping results. I don't understand why NetWorker has to
> > > clone
> > > > the last two savesets to another volume member when there's plenty
> of
> > > > space on the first one, and it started on the first one. Makes no
> > > sense,
> > > > and like I said, the results change back and forth with every run.
> I
> > > > even tried cloning just two savesets, one on one tape (ARC001) and
> one
> > > > index on another (ARC002) and same results. This does not occur
> when
> > > > both ssids are on the same tape, HOWEVER.
> > > >
> > > > I've cloned lots of stuff before to other clone pools (not ARC
> Clone),
> > > > and I've never seen this behavior, but I never had more than one
> > > > available tape before either. Hmm.... My experience was always
> been
> > > that
> > > > NetWorker didn't care how many ssids were contained on how many
> > > > different tapes. It would clone all of them to the same clone
> volume
> > > and
> > > > not to a different one until it ran out of space on the former
> then it
> > > > goes to the next.
> > > >
> > > > Anyone seen this? I tried re-labeling the tapes, still happens.
> We're
> > > > runnng 6.1.1. Storage node is using Linux Red Hat, and primary
> server
> > > is
> > > > running Solaris 2.8. also 6.1.1. If I delete one of the clone
> volumes,
> > > > so I only have one, and try again, it just sits and waits for the
> > > second
> > > > volume after cloning the first ssid(s). Not good.
> > > >
> > > > Noticed that if I re-label one of the two ARC Clone volumes as
> > > something
> > > > else -- any other pool -- so I only have one ARC Clone pool tape
> > > > available, NetWorker acts normal. It writes everything to just the
> one
> > > > tape and never issues any device or resource messages. For some
> > > reason,
> > > > though, having two available clone volumes causes the problem, and
> I
> > > > have the sessions for each device set to 5, and the number I'm
> cloning
> > > > is less than that. Strange! The thing is that I would like to
> have
> > > two
> > > > clone volumes available so
> > > > when one fills up, NetWorker will have another one to load. I
> can't
> > > rely
> > > > on auto-media management because NetWorker will not necessarily
> select
> > > a
> > > > tape that has the proper bar code.
> > > >
> > > > Would greatly appreciate any help.
> > > >
> > > > Thanks.
> > > >
> > > > George
> > > > George.Sinclair AT noaa DOT gov
> > > >
> > > > --
> > > > Note: To sign off this list, send a "signoff networker" command
> via
> > > email
> > > > to listserv AT listmail.temple DOT edu or visit the list's Web site at
> > > > http://listmail.temple.edu/archives/networker.html where you can
> > > > also view and post messages to the list.
> > > > =*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
> > >
> > >
> >
> ************************************************************************
> **************************
> > > The contents of this email and any attachments are confidential.
> > > It is intended for the named recipient(s) only.
> > > If you have received this email in error please notify the system
> manager or the
> > > sender immediately and do not disclose the contents to any one or
> make copies.
> > >
> > > MBI - System Team
> > >
> >
> ************************************************************************
> **************************
> > >
> > > --
> > > Note: To sign off this list, send a "signoff networker" command via
> email
> > > to listserv AT listmail.temple DOT edu or visit the list's Web site at
> > > http://listmail.temple.edu/archives/networker.html where you can
> > > also view and post messages to the list.
> > > =*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
>
> **************************************************************************************************
> The contents of this email and any attachments are confidential.
> It is intended for the named recipient(s) only.
> If you have received this email in error please notify the system manager or
> the
> sender immediately and do not disclose the contents to any one or make copies.
>
> MBI - System Team
> **************************************************************************************************
--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
|
