in the java adm console help

scan is drive is sso but no available scan host......

"E Wensell" wrote:
>
>I have upgraded my master and shared-storage media servers from NetBackup
>5.1 to 6.0MP3. Several, not all, of my drives are coming up SCAN-TLD in
Device
>Monitor.
>
>What is SCAN-TLD? I have not found any mention of it in the docs, knowledge
>base, or in Google.

"Chan" wrote:
>
>in the java adm console help
>
>scan is drive is sso but no available scan host......

Thank you Chan, didn't realize it was the same thing. For starters, my drives
all appear to be operational. Even though they say "SCAN-TLD", hosts are
still able to allocate them for backups. None of my drives appear "DOWN".

On the surface it seems quite benign, but even so any suggestions on how
to correct it?

I can take a stab at it. I'd venture to guess that your media servers are
lying to you.
I'd shut down most of the media servers except one and work your way back.
I've seen drives showing good status but when it comes to using it, it
will be downed.


"E Wensell" wrote:
>
>"Chan" wrote:
>>
>>in the java adm console help
>>
>>scan is drive is sso but no available scan host......
>
>Thank you Chan, didn't realize it was the same thing. For starters, my drives
>all appear to be operational. Even though they say "SCAN-TLD", hosts are
>still able to allocate them for backups. None of my drives appear "DOWN".
>
>On the surface it seems quite benign, but even so any suggestions on how
>to correct it?

"Chan" wrote:
>
>I can take a stab at it. I'd venture to guess that your media servers are
>lying to you.
>I'd shut down most of the media servers except one and work your way back.
> I've seen drives showing good status but when it comes to using it, it
>will be downed.
>
>
>"E Wensell" wrote:
>>
>>"Chan" wrote:
>>>
>>>in the java adm console help
>>>
>>>scan is drive is sso but no available scan host......
>>
>>Thank you Chan, didn't realize it was the same thing. For starters, my
drives
>>all appear to be operational. Even though they say "SCAN-TLD", hosts are
>>still able to allocate them for backups. None of my drives appear "DOWN".
>>
>>On the surface it seems quite benign, but even so any suggestions on how
>>to correct it?
>

I have the same issue - after doing some troubleshooting, it looks like one
of my media server (the master in this case) has an issue accessing one of
my sso drives. I have eventlog errors referring to problems with one specific
drive:
Event id 4111, Source NetBackup AVR Daemon, GetTapeParameters(0) failed on
drive .... (device 2,\\.\Tape2) The requested resource is in use.

Chan wrote:
> I can take a stab at it. I'd venture to guess that your media servers are
> lying to you.
> I'd shut down most of the media servers except one and work your way back.
> I've seen drives showing good status but when it comes to using it, it
> will be downed.
>
>
> "E Wensell" wrote:
>> "Chan" wrote:
>>> in the java adm console help
>>>
>>> scan is drive is sso but no available scan host......
>> Thank you Chan, didn't realize it was the same thing. For starters, my drives
>> all appear to be operational. Even though they say "SCAN-TLD", hosts are
>> still able to allocate them for backups. None of my drives appear "DOWN".
>>
>> On the surface it seems quite benign, but even so any suggestions on how
>> to correct it?
>

Hi All,

Look at the vm.conf parameter; "SSO_SCAN_ABILITY = [0 - 9]"
Where 0 is NO ABILITY, 1 is LOWEST, and 9 is HIGHEST.

Letting your SSO hosts work out who is te SCAN host for themselves is
fraught with danger, mostly due to the election process for the SCAN
host - you can end up with multiple hosts performing the SCAN function
across the SSO, which in the end just doesn't work. Forcing the levels
using this parameter will take the guesswork out of maintaining a
working SSO configuration and minimize Error 52's in your jobs and other
tape mount issues.

I'd set your Master to use level 9; SSO_SCAN_ABILITY = 9
(Your Master does have access to ALL drives - yes?)

Set any robot control hosts to level 5; SSO_SCAN_ABILITY = 5


Set any SAN Media servers or Media servers without robot control to level 0.

After doing something like the above, restart your NetBackup services on
all MM's.


To verify the existing SSO configuration and levels, use the command ;
/volmgr/bin/vmdareq on the Master server.

vmdareq without args will give you the status of all SSO drives, which
Media Servers are registered, which have them reserved (in use), and
finally who is the SCAN_HOST for a particular drive. The "-driveinfo"
switch to this command, gives this output too.

The other form of the command gives you a summary table showing the
SCAN_ABILITY of each host in the SSO configuration;

vmdareq -hostinfo -a

I'll say (no offence) - RTFM for this command to get further information
on it's arguments, capabilities and ramifications. Like vmoprcmd, it is
not just a status display utility but can change your running SSO
configuration. You'll also need to read further if your Master server
isn't your Volume DB host (-h ). Or you wwant to run this
command on one of your Media Servers, (-M ).

Hope this helps, or if you require further assistance, drop us a line.

Regards,
Andrew R-C.

Andrew Ross-Costello wrote:

>Hi All,
>
>Look at the vm.conf parameter; "SSO_SCAN_ABILITY = [0 - 9]"
>Where 0 is NO ABILITY, 1 is LOWEST, and 9 is HIGHEST.

Thank you for your response. It did not work for me, but got me going in
the right direction... Mostly...

I started out adding SSO_SCAN_ABILITY = x to vm.conf on the master (9) and
all the media servers (5). Restarted the medias and masters, and "vmdareq
-hostinfo -a" still reported all hosts has having a "5".

More Googling and I found http://seer.entsupport.symantec.com/docs/278996.htm
. It basically says vm.conf is obsolete and to use nbemmcmd instead. So I
did:

nbemmcmd -changesetting -sso_scan_ability X -machinename

Setting the master to 9 and all the other hosts to 0. Restarted the master
and all media servers. The value shows correctly when using "nbemmcmd -listsettings"
for each host, but "vmdareq -hostinfo -a" still shows all hosts have a 5,
and am still seeing random-like SCAN-TLD under the Device Monitor in the
Admin Console.

Again, SCAN-TLD appears to be benign. I have had no errors caused by it.

>I'll say (no offence) - RTFM for this command to get further information

>on it's arguments, capabilities and ramifications.

I hate to ask this, but can you point out which 6.0 document it is listed
in. I have been unable to find vmdareq in my doc sets, nor are there any
man pages for it on my systems. I am under the impression vmdareq is an obsolete
(or soon to be) command.

It all appears the 5.x "SSO Installation Guide" has been merged into Media
Manager Admin Guide.

Thanks again for your responses...

"E Wensell" wrote:
>
>I started out adding SSO_SCAN_ABILITY = x to vm.conf on the master (9)
and
>all the media servers (5). Restarted the medias and masters, and "vmdareq
>-hostinfo -a" still reported all hosts has having a "5".
>

Typo... media servers (0)

"E Wensell" wrote:
>
>I have upgraded my master and shared-storage media servers from NetBackup
>5.1 to 6.0MP3. Several, not all, of my drives are coming up SCAN-TLD in
Device
>Monitor.

I'm beginning to really think is is nothing.

Further clarification, I am seeing this in the BOTTOM panel of Device Monitor.
In the top panel, all hosts show TLD. If I select a drive, all the hosts
which have access to that drive are listed in the bottom panel. At least
one host for each drive shows SCAN-TLD.

If I go to the command line and run "vmdareq -display -D DriveX", SCAN_HOST
is displayed next to the same host which has SCAN-TLD in the GUI...

So, my guess is the GUI is simply telling me which is the current scan host.
It is just something I had never seen in 5.1 .

E Wensell wrote:
> "E Wensell" wrote:
>> I have upgraded my master and shared-storage media servers from NetBackup
>> 5.1 to 6.0MP3. Several, not all, of my drives are coming up SCAN-TLD in
> Device
>> Monitor.
>
> I'm beginning to really think is is nothing.
>
> Further clarification, I am seeing this in the BOTTOM panel of Device Monitor.
> In the top panel, all hosts show TLD. If I select a drive, all the hosts
> which have access to that drive are listed in the bottom panel. At least
> one host for each drive shows SCAN-TLD.
>
> If I go to the command line and run "vmdareq -display -D DriveX", SCAN_HOST
> is displayed next to the same host which has SCAN-TLD in the GUI...
>
> So, my guess is the GUI is simply telling me which is the current scan host.
> It is just something I had never seen in 5.1 .

I have no pointers to vmdareq in the NB6.0 documentation, it (the
command/binary) is there in the distribution at the moment, but the
nbemmcmd seems to be the way to go, as you've found out.

Apologies for my red-herrings, my SSO configuration is under NB5.1 and
my exposure to NB6.0 is in a single library with one drive
configuration, so I've yet to experience SSO under NB6.0, although your
messages have been a great tutorial in what to expect. And based on your
last message above, I would concur in that the GUI is now trying to
inform you on which host is the SCAN_HOST for each drive.


Regards,
Andrew.

sorry for offtopic but this is the one and only place in internet when this problem has been touched

One update for you:

in NBU 6.0 - 6.5 dont use vmdareq.

Do that way :
nbemmcmd -changesetting -sso_scan_ability X -machinename
restart nbu services on media and master, and check for results:
nbemmcmd -listsettings -machinename

look for : SSO_SCAN_ABILITY="X"

btw check this one also: http://seer.support.veritas.com/docs/269656.htm

and no more 84 errors
cheers !