On 25/06/12 13:11, Div Prakash wrote:
> Hi,
> 
> I think your method describe two steps:
> 
> 1. Stage all the ssid to a new pool on old server
> 2. To scan the tapes on to another server with pool having staged data.
> 

Thierry's suggestion is good in many ways. The advantage is that you
potentially only need to scan in a fraction of the number of tapes.
Scanning is an onerous manual process.

However the staging process may or may not be straightforward. If you
already have all or most of your tapes in your library then you can
simply create a file of all the SSIDs you need to scan and run a single
nsrstage command. It might take several days but it should be automatic.
However if most of your tapes are offsite it adds some complications and
might make the idea less useful.

On the whole though, I like the idea and think it is worthwhile.


> There is another way to do it I think;
> 
> 1. just create the same client resources(same client id), same
> pool,group,schedule,barcode,policy etc.
> 2. use scanner -i client name device name (after loading the tape having
> the client data)
> 
> may be I should just copy the client index folder directly on the server
> and run only #scanner -m -c clientname device name. Bit time can be save to
> just fetch savests/vol records to media database only rather than building
> index for each file.

This might save a little time, but is just as much work.

> 
> I am not sure which is going to faster way, i think it will depend the
> number of clients data will be having on the required tapes and also total
> number of tapes for the client data. we have three clients not having more
> than 9 tapes for each client. But there is one storage node having huge
> number of tapes.
> 
> Thanks
> 
> 
> On Mon, Jun 25, 2012 at 5:23 PM, FAIDHERBE Thierry <
> thierry.faidherbe AT forem DOT be> wrote:
> 
>>  To perform tape scan, to must define same pool on target networker
>> server.
>> So, ...
>>
>> For a faster "import" and clean import, you could, on the old
>> AND new server, create a new pool. Ex TRANSFER. Same name on both env.
>>
>>
>> Then, for all the clients you need to move from the old to the new
>> env :
>> 1) on the new env, define old clients with same clientid.
>>
>> 2) in a reverse time order (newest backup first), stage backups
>> from these clients, from the current old pool to the newly created
>> TRANSFER pool on the old env.
>> Then, when tape is full, for sample, on the new env, use "scanner -m"
>> to let it recognize the tape and its content.
>> (it will only be clients you need to move, tape not filled in with unneeded
>> data. So less tape to scan).
>>
>> These newly scanned backups will be recoverable (meaning you can recover
>> from,
>> based on ssid or file based if you know complete path).
>>
>> At this stage, 2 options for you:
>> A) if you absolutely need to rebuild CFI, then scanner -I on
>> each SSID. As nsrstage moved 1 ssid a time, your rebuilt will be faster
>> as each tape block scanner will read belongs to same ssid. (no tape
>> multiplex)
>>
>> B) you can, from the old nsr server, issue an nsrinfo cmd to gather
>> file list from the old CFIs. Store them in a text file and keep them
>> on the new nsr server in case of need.
>>
>> HTH
>>
>> Thierry
>>
>>
>> -----Message d'origine-----
>> De : EMC NetWorker discussion [mailto:NETWORKER AT LISTSERV.TEMPLE DOT EDU] 
>> De la
>> part de Div Prakash
>> Envoyé : lundi 25 juin 2012 13:20
>> À : NETWORKER AT LISTSERV.TEMPLE DOT EDU
>> Objet : Re: [Networker] Migrating few critical clients to a diffrent
>> networker
>> serve
>>
>> Hi Rovin,
>>
>> We cant do the server migration. We have a old networker server having few
>> critical clients running on it. Customer wants it to decommission.
>> We need to move those clients to already existing networker server (having
>> 300+ clients already running on it) with there existing data. So will not
>> be able to overwrite the media database.
>>
>> I also got the confirmation from our Asia pacific expert, scan of the all
>> the tapes is only option to perform this task.
>>
>> Thanks
>>
>>
>>
>> On Mon, Jun 25, 2012 at 3:12 PM, Dsouza, Rovinabishek <
>> rovinabishek.dsouza AT hp DOT com> wrote:
>>
>>> How about following the NetWorker Server Migration ?  You will all the
>>> data that you wanted and that too without much complications.
>>>
>>> Rovin A. D'Souza
>>> rovinabi AT gmail DOT com
>>>
>>> -----Original Message-----
>>> From: EMC NetWorker discussion [mailto:NETWORKER AT LISTSERV.TEMPLE DOT 
>>> EDU] On
>>> Behalf Of bingo
>>> Sent: Monday, June 25, 2012 3:18 AM
>>> To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
>>> Subject: [Networker] Migrating few critical clients to a diffrent
>>> networker serve
>>>
>>>  The problem is the media index. The only way to add existing but unknown
>>> save sets is to scan the backup media. Of course, nobody wants to do
>> this.
>>>
>>> So the easiest way would simply to keep the old server for recoveries.
>>> Otherwise the task could keep you busy for a while.
>>>
>>> +----------------------------------------------------------------------
>>> |This was sent by carsten_reinfeld AT avus-cr DOT de via Backup Central.
>>> |Forward SPAM to abuse AT backupcentral DOT com.
>>> +----------------------------------------------------------------------
>>>
>>