New client

[gudguy]

Hi everyone.

I'm doing a directed recover of Exchange server which is not on client list. So I have added the client list and ran scanner command to restore client database. After it is done, when I tried to restore the msg pops saying "no index was found..." I ran nsrck -c and found that this new server has 0 records occupying 0 KB. My question is where did the restored client database is? How do I make it work?

Thanks in advance.

 

[wallace88]

Running nsrck -c after the scanner command simply erased all the index data and recreated a blank index.

Are you sure you got the correct client name?

Load the tape in the drive, and then run scanner -v to get a list if savesets that are on the tape. Once you know what's on the tape, then cerate the appropriate client.

 

[gudguy]

Yes I'm sure the client name is correct. Actually even before running nsrck -c, the records were 0. I ran nsrck -c after running scanner command for the second time, since I noticed the records were 0 when I first ran scanner command. Well...I'll run it again and see.

 

[wallace88]

How the the client name show up from the scanner -v output, and from the NetWorker administrator program? Are they the same format (i.e, shot name or fully qualified domain name)?

Does the client resource has the alias defined with both short and FQDN?

When you are scanning in the data, use scanner -iv so you can see what is being read from the tape while it is scaning.

 

[gudguy]

First of all thanks for your valuable input Wallace. Actually I'm not at my work place now. But when we first started the scan, this client name isn't there in Networker's administrator program. I added it in Networker's program and restarted the scan. Still I could see no files under \usr\index\servername\db6 on networker server (on unix). Thou there are no files, I ran the 'nsrck -c servername' and resarted the scan, this time not giving SSID. I've got SSID from our old mminfo files. Btw, the networker server I'm using to restore is not the original server.

My only question is which directory were the client db files restoring to when we scanned the server twice? Do you think running scanner command without SSID will help? If yes, I'll try tomorrow when I goto work.

 

[wallace88]

\usr\index\servername\db6?? This is not the default index path in Unix. Usually this is /nsr/index/<client name>

If you searched your file system with:
find / -name <client name> -print
how many times does it show up? If there are more that one, is the index database there?

You did have to create the client before scanning. Hwever, I'm wondering if the index path for this client on the original backup server is different from what you set it as now on this server.

Like I said, use scanner -iv, and capture the output so that you know what scanner is doing.

 

[gudguy]

I'm sorry. I meant \nsr\ not \usr\. Secondly the default path in latest legato version has been replaced with \nsr\index\<clientname>\db6\.

When I do scanning for other servers, the files are stored into default path. Let me scan again and give you more input.

 

[gudguy]

I ran scanner -iv /dev/rmt/2cbn and it finished 1st tape. Its now asking for 2nd tape. Before inserting second tape, I went and checked \nsr\index\<clientname>\db6\, and found no files except for default header files. I was wondering whether I should continue to scan with 2nd tape.

 

[wallace88]

Same question... where is the index information being saved, since it doesn't appear to be saving in the default location? This is what you need to find out.

 

[gudguy]

Since I'm new to legato I got stuck up there Wallace. Can I find out in networker admin? I mean if all other client servers' indexes are being restored respective default paths, how can I find out this particular <client name>'s default path?

 

[gudguy]

There are 9 exchange servers that are backed up onto these tapes, one of which is the server I'm trying to restore. Since the scan is not completed, it could be that it restored the client indexes of other servers first. So what I'll do is, I'll stop the present scan and run scan again with specific SSID that I know of this clientname and will post the info.

 

[gudguy]

Wallace,

I started the scan with SSID: 907684865 and filenumber: 6.

Its scanning but in between I saw an error msg: I paused it to get the screen shot which is:

Scanner: scanning file 6, record 1600
Scanner: (ssid 907684865) Can't connect to index service on phccshop-dev07
Scanner: RPC error: Program not registered
Scanner: write failed, Bad file number
Scanner: scanning file 6, record 1700

Does it mean anything? Can I just scan without file number?

 

[wallace88]

The scanner error is a symptom why you don't have any index information. It could be caused by a number of reasons:

1. The tape is damaged

2. The o/s cannot read the tape properly. This can happen if the tape was written in one environment, and being read from another environment with the driver settings different.

3. backups running while you are scanning in the tape

4. You could be experiencing a software bug.

I would verify that the tape can be read on the original server. If not possible, then I would try to verify that the tape device has the same o/s driver settings and block size settings. In other words, is the current backup server setup in the same way as the original backup server in terms of tape driver, controller.

If you can't figure this out, then I suggest you open a support ticket with Legato so that they can help you analize your scanner problems. The problem you're seeing may need more in depth analysis.