Networker

Re: [Networker] Multiple logical adv_file devices + single physical disk device?

2007-10-01 09:48:46
Subject: Re: [Networker] Multiple logical adv_file devices + single physical disk device?
From: Davina Treiber <Davina.Treiber AT PEEVRO.CO DOT UK>
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Date: Mon, 1 Oct 2007 14:41:59 +0100
Lee Roth wrote:
Networker for Unix V7.4 (Solaris)

Currently, I have a 1TB disk setup that is mounted as an adv_file device.
I have a single media pool called "DiskBackup" that uses the physical
adv_file device (e.g. mount is "/large_disk"). Backups occur to disk, the
savesets are then staged to tape based on fullness of the disk device.

Networker, of course, treats this 1TB of disk space as two volumes: The
"normal" one that it writes to and a "read-only (RO)" twin that can be
used (in some situations) for concurrent operations.

On occasion, the RO part of the disk space gets tied up with some
long-winded operation and isn't free for other tasks that might need only
comparatively brief read-only access.

Would doing the following be possible/good/bad... (am I insane)?:

1) Create a one-per-client symbolic link to the physical 1TB disk device, e.g

  ln -s /large_disk /client_disk/alpha
  ln -s /large_disk /client_disk/bravo
  ln -s /large_disk /client_disk/charlie
  ...

2) Create a unique media pool for each client that points to its own
personal symbolic name:

  Pool:  backup_alpha    Target Device:  /client_disk/alpha
  Pool:  backup_bravo    Target Device:  /client_disk/bravo
  Pool:  backup_charlie  Target Device:  /client_disk/charlie

Make sure that each client-specific pool specified their client as the
one-and-only client that uses that pool.

---

My thinking is that if Networker is busy cloning client "alpha" savesets
to tape drive #1, then I could start another clone operation of client
"bravo" savesets (to a different tape drive, media from different tape
pool) that would not have to wait for the RO disk device to first become
free from its long-winded cloning operations with client "alpha" savesets.

Obviously, if only reading is going on no real worries, but would horrible
things happen if simultaneous backups of these clients were going on?

Thanks,


Doesn't sound like a very manageable solution when you go above a handful of clients. Would probably work but is not very scalable. Also bear in mind that all the devices will become full at the same time.

To sign off this list, send email to listserv AT listserv.temple DOT edu and type 
"signoff networker" in the body of the email. Please write to networker-request 
AT listserv.temple DOT edu if you have any problems with this list. You can access the 
archives at http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER

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