TSM v ADSM

[benny_01670]

Hi all,



I'm new to this site, so firstly hello all.



I'm currently looking into upgrading our ADSM (3.1) to the latest TSM (5.?).



What I'm trying to find is information on what TSM would give us over ADSM. What will it improve on, what it doesn't improve on etc etc.



Basically what are the benefits of moving to TSM.



I know one of the major benefits would be Support from IBM. But that alone will not be enough to convince the management that this upgrade is required.



Many thanks in advance

Ben Swinney

 

[n9kws]

There are a number of reason to upgrade to the latest version of TSM, and support by IBM should not be dismissed so easily. In no particular order, here are some of the new features in TSM products:



1) Support for new devices -- Including tape drives, libraries, and fiber connectivity.



2) Lan-Free backups -- The ability to have backup go striaght to a device with passing through the TSM server.



3) More & Better multi-threading capabilities -- multi-threaded backups, are restores,



4) Support for Data Protection agents which works.



5) Detailed Perfromance Monitoring & Reporting in a graphical interface.



6) Integration into the Tivoli mnagement framework.



7) Commands to help you manage your storage pools better.



Support for the current crop of desktop OS's and most of the popular server os's.



9) WEB based admin & client interfaces in additon to the GUI based clients.



10) Lots of redbooks covering new features, and implemetation information.



11) Rumor Alert: I am told that in V5.2 will be collocation groups. The ability to declare a group of clients as colloc from with within an existing storage pool which is NOT currently designated as a collocation storage pool.



Much of the basic functionallity continues but these improvements listed above, and probably more that I have not named, make the current TSM a much improved product over it's eariler versions.



Lastly, I cannot stress enough the imporatnce of IBM support. Especially, when you at DR testing and discover a problem.

 

[LED888]

Ben,



Changes in Defining Drives and Libraries

Device special file names and External Library managers are now specified in the DEFINE PATH and UPDATE PATH commands, rather than in the DEFINE DRIVE, UPDATE DRIVE, DEFINE LIBRARY, and UPDATE LIBRARY commands.



See Adding a Manual Library, Adding an Automated Tape Library and Tivoli(R) Storage Manager Administrator's Reference.





Moving Data by Node

You can use the MOVE NODEDATA command to move data in a sequential-access storage pool for one or more nodes, or move selected file spaces for a single node. You can also use MOVE NODEDATA to move data to another storage pool.

See Tivoli Storage Manager Administrator's Reference.





Support for Simultaneous Writes to Primary and Copy Storage Pools

You can specify copy storage pools in a primary storage pool definition. When a client backs up, archives, or migrates a file, the file is written to the primary storage pool and is simultaneously stored into each copy storage pool.

See Tivoli Storage Manager Administrator's Reference.





High Availability Cluster Multiprocessing

Tivoli Storage Manager can now use High Availability Cluster Multiprocessing (HACMP). HACMP provides the leading AIX-based clustering solution, which allows automatic system recovery during system failure detection. By using HACMP together with Tivoli Storage Manger, you can ensure server availability.



Tivoli Data Protection for New Network Data Management Protocol Support

New Network Data Management Protocol (NDMP) support now extends to the AIX (32-bit and 64-bit) Tivoli Storage Manager server platform. The new Tivoli Data Protection for NDMP product supports NDMP backup and restore for network-attached storage (NAS) file servers from Network Appliance. NDMP allows a network storage-management application to control the backup and restore of an NDMP-compliant file server without installing third-party software on that server. The NAS file server does not require installation of TSM software. The TSM server uses NDMP to connect to the NAS file server to initiate, control, and monitor a file system backup or restore operation. The NDMP support for NAS file servers enables higher performance backup to tape devices without moving the data over the LAN. TDP for NDMP is a separately priced and licensed product.



See Tivoli Storage Manager Administrator's Reference.





Data Validation with Cyclic Redundancy Checking

Tivoli Storage Manager provides the option of specifying whether a cyclic redundancy check (CRC) is performed during a client session with the server, or for storage pools. The server validates the data by using a cyclic redundancy check which can help identify data corruption. Data validation can be enabled for one or all of the following:

Tivoli Storage Manager client nodes at Version 5.1.

Tivoli Storage Manager storage agents at Version 5.1.

Storage pools

See Tivoli Storage Manager Managed System for SAN Storage Agent User's Guide and Tivoli Storage Manager Administrator's Guidefor more information.





New Licensing Method

The new licensing method enables you to register the exact number of licenses that are required, rather than in increments of 1, 5, 10, and 50.

See Licensing Wizard and Tivoli Storage Manager Administrator's Reference.





Server Performance Enhancements

There are two new Tivoli Storage Manager performance enhancements:

AIX Asynchronous I/O Support. This feature is available via a new option in the server options file.

AIX Direct I/O Support. This feature is available via a new option in the server options file.

See Tivoli Storage Manager Administrator's Reference.





Here is the URL for the online manuals:

http://publib.boulder.ibm.com/tividd/td/tdmktlist.html



Since you are aware that ADSM 3.1 is no longer supported.

If you find a bug or a problem with the code. The only thing that you can do is to install

the latest maintence. If that is already installed there is really nothing that you can do since IBM/Tivoli is not developing any maintence code for ADSM 3.1 .

I would not install the latest patch code since this code have not been fully tested like the maintence code. With the patch code you can solve one issue and possibly introduce another unknown issue.



Sias

 

[benny_01670]

Cheers guys.



Will add those things into my report. It's quite difficult to find information on TSM v ADSM.



Thanks once again.



Ben Swinney

 

[benny_01670]

I have one last question to ask.



As the majority of you have used /using TSM.



Could you please tell me what the shortcommings of it are.



e.g. Whats are it's issues, what are it's problems, what could be improved etc.



Many thanks in advance



Ben Swinney

 

[n9kws]

<TABLE BORDER=0 ALIGN=CENTER WIDTH=85%><TR><TD><font class="pn-sub">Quote:</font><HR></TD></TR><TR><TD><FONT class="pn-sub"><BLOCKQUOTE>

Could you please tell me what the shortcommings of it are.

</BLOCKQUOTE></FONT></TD></TR><TR><TD><HR></TD></TR></TABLE>



Mostly, My biggest problem with TSM is getting the PHB to understand how it works!



Seriously, It does everything I need to to do and does it pretty well. Sometimes my PHB complains about the costs associated with it. Overall, me and my company are quite happy with it.

 

[bensnyder]

My biggest problem is debatable, IMO, on whether or not it's TSM's fault.



The filesystem that our customer data lives on has tons of harlinks (each customer dir has well over 1000 files that are hardlinked). TSM seems to not realize these files are hardlinked until it's migrating from the diskpool to the tapepool. Plus, in order for hardlinks to be restored properly, EVERY instance of the linked file needs to be restored (a restore of a single customer requires a manual re-linking).



The result is that during a backup I am sending ~400,000 extra files per server over the network, which causes my backups, and restores, to take an extremely long time. Even on an incremental, each instance of the file is backed up because if a customer is added or deleted from the server, the modify time on the inode changes. I know from experience other backup products handle hardlinks a little better because the transport mechanism used natively supports them.



We are investigating creating a script to handle these linked files for us (an automated/manual re-linking) to facilitate an exclude on the directories/files in question.

 

[LED888]

<TABLE BORDER=0 ALIGN=CENTER WIDTH=85%><TR><TD><font class="pn-sub">Quote:</font><HR></TD></TR><TR><TD><FONT class="pn-sub"><BLOCKQUOTE>



Could you please tell me what the shortcommings of it are.



Whats are it's issues, what are it's problems, what could be improved etc.



</BLOCKQUOTE></FONT></TD></TR><TR><TD><HR></TD></TR></TABLE>



There are some System Admin (may be some TSM Sales Rep) believes that TSM can backup and restore the operating system without any problem.



With os there are a lot of open files were the information is constantly changing. When TSM backup an open file, if the os allows it, the file will be backed up what the file looked like at that moment. This is known in the TSM world as a "fuzzy backup."



When the file is restored, the appliaction may or may not crash. Depending on the file(s).



Had a System Admin that backed up his Solaris OS.

After a TSM restore, the os keep on crashing on him.

He could not figure out why.



Windows os have there files locked.

Since the file(s) are locked by the os, TSM can not backup the file(s).

You can force TSM to backup the file, due to the lock the file will not be backed up.



For AIX, if you want to backup the os. I would do a mksysb or a savevg. I would not use TSM. There is a product from IBM called Sysback. It really makes backing up and restoring AIX simple.





Just my 2 cents.



Sias