I have had this happen to me. The file system log corrupted and the disk
was returning I/O errors, so there were partial images on the DSSU, but
it purged the associated catalog data when the backup failed.
Try the bpexpdate to expire the images first. If NetBackup tells you the
image doesn't
Besides the missing libstdc++-compat libraries, also check to see if you
have SELinux set to Enforcing mode and if there's SELinux errors being
logged. If so, you might have to set it to Permissive mode.
Eric
-Original Message-
From: seanharv [mailto:[EMAIL PROTECTED]
Sent: Tuesday,
I sync'd our production NetBackup 5.1MP6 installation to a test server
we have (normally used for DR) and upgraded that to 6.5, then patched to
6.5.1. Everything worked as documented, no problems were encountered
other than one inconsitency that NBCC caught (there was a frozen tape
with expired
I believe this is only possible if you're using NBAC/VxSS with each
operator logging in with a different username, and then enabling logging
for bpjava-msvc. Even then, you have to trace back in the log a bit to
see the user. It's sort of a nuisance.
If anybody knows a better way, I'd like to
NetBackup 6.5.1 master server on Solaris 10 update 4. System uses LDAP
to authenticate logins vs. Active Directory on Win2k3 R2. This works
fine for local and SSH logins. Using it to authenticate NBAC logins
doesn't seem to work though.
The root broker is running on the master server and is
On Windows, this info is in the registry:
HKEY_CURRENT_USER\Software\JavaSoft\Prefs\vrts\nbe\/N/B/E/Login/Lists\vr
ts-nbe-/Admin/Console
(yes, they actually used forward slashes in the names..) There is a key
called host/List that contains the list.
Eric
In planning our upgrade to NB6.5, I've also been looking at the NBAC
feature. Is anyone here using it in it's complete implementation (ie,
including client authentication) and have any comments about it?
Is full client auth overkill? It seems like the certificates that are
issued via the
Hi Bobby,
I believe they are telling you the truth. I've set up the MS SQL agent
as well, and every database comes through as it's own backup job. I'm
not aware of any way to alter this behaviour, but I haven't really
looked into it.
Eric
From: Bobby
How do most people back up SAN data with NetBackup? Currently, we have a
legacy setup where the NB client on each SAN-connected host system backs
up the SAN mounts that it sees as a regular file system. Obviously, this
is very inefficient as it puts load on each client and requires
additional
I've noticed lately that when we perform our test restores, they
frequently complete with a status 1 (Partially successful) and the
warning of Restore must be resumed prior to first image expiration on
DATE. There doesn't appear to be any other error on exit either in the
GUI or in the error log,
Hello,
Our current libraries (a Sun L25 and a Sun L700), are configured so that
they do NOT show the Media Type suffix when NetBackup 5.1MP6 does an
inventory. So a tape with a physical label of AB0123L2 shows as
AB0123 and no NetBackup media ID generation rules are required.
In a DR situation,
Subject: Re: [Veritas-bu] Backing up ZFS snapshots
Johnson, Eric schrieb:
You can control whether the .zfs directory for snapshots is visible or
not woth the zfs set snapdir=visible filesyste option.
I know - but how does that solve the problem? If .zfs were
visible, wouldn't this make even more
You can control whether the .zfs directory for snapshots is visible or
not woth the zfs set snapdir=visible filesyste option. From the zfs
manpage:
snapdir=hidden | visible
Controls whether the .zfs directory is hidden or visi-
ble in the root of the file system as
I get 20-30MB/sec with the inline copy from our DSSU. Without the inline
copy, the speed is 40-50MB/sec. So yes, we see the 50% speed hit. Not
sure why.
Eric
From: Santos, Tarso D [mailto:[EMAIL PROTECTED]
Sent: Tuesday, March 06, 2007 12:49 AM
To:
Turns out the fix for the hanging backups we were experiencing with the
latest Solaris 10 patch cluster with NBU5.1MP6 were indeed related to
the ip:do_tcp_fusion kernel parameter. Although Sun claims this bug was
fixed in 118833-24, it still exists in 118833-36. The suggestion was to
turn it off
We've installed the latest Solaris 10 patch cluster (we were at kernel
118822-24 before I believe, we're now at 118833-36). Since then, backups
that go from network to tape will randomly freeze up and stop writing
data, although the job sits there and looks active. Sometimes it will
take a few
You can also look at the 'bperror' output, as it will show the speed for
each fragment was copied to tape. It's rather inconvenient though.
Eric
-Original Message-
From: Martin, Jonathan (Contractor) [mailto:[EMAIL PROTECTED]
Sent: Wednesday, January 03, 2007 8:57 AM
To: NB List Mail
Has anyone used the ZFS file system with NetBackup 5.1 and can comment
on any issues or gotchas?
We'd like to move our catalog and two DSSU file systems from UFS to ZFS.
Does ZFS report a full file system status in a way that NetBackup can
understand so that it will expire old images on the DSSU
In planning for disaster recovery, I've been examining different situations
that aren't a complete loss of the NetBackup system. One is where our DSSU unit
is lost (file system corrupted beyond repair, or whatever). Obviously, this
leaves you with no images in a location that NetBackup thinks
We are in the early stages of migrating from Legato NetWorker to Veritas
NetBackup 5.1. As part of our migration, we are adding enough disk space
to hold one week worth of backups before they are staged to tape.
When configuring our disk (approximately 9.6TB before RAID in a Sun
StorEdge 3511
20 matches
Mail list logo
