Technical Support Services Blog
Discover the latest trends in technology, and the technical issues customers are overcoming with the aid of HP Technology Services.

DPTIPS: Device Block Size Challenges

Alrighty then, it's time to integrate your shiny new HP D2D StoreOnce appliance into your Data Protector environment.  You're an old hand at DP, so that's no issue.  For the new D2D, you've read the Concepts Guide, Install & Config Guide, User Guide, and finally the all-important HP D2D Backup Systems Best practices for VTL, NAS and Replication implementations document.  Among all the data swimming in your head, you remember reading on page 25 of that last tome ...

Use a block size of 256KB or greater. For HP Data Protector and EMC Networker software a block size of 512 KB has been found to provide the best deduplication ratio and performance balance.

Not a problem.  You know the default for LTO/Ultrium is 64k, and you've typically increased that to 256k which has historically been the sweet spot for backup to physical tape.  So now you're going to set your VTL tape device block size to 512k.  What could possibly go wrong?

 

Well, a couple of things, specifically.  If your Media Agent (MA) host is running MS-Windows or HP-UX, you may encounter a message similar to the following.

 

Size_Error_Msg.jpg

 

Media Agent host is HP-UX

If you are running HP-UX 11i v2 (11.23) or earlier, the stape driver is limited to a maximum block size of 256k.  Attempting the use of a larger block size until recently would work for the DP write operation but fail miserably on restores.  This was caused by the stape driver quietly "chunking" larger block writes down to multiple 256k block writes.  Details are included in Document ID KM915678 and Document ID c02256887.

 

Don't fret if you're stuck at 11i v2 which dictates a max block size of 256k.  It's infinitely better than the default of 64k where you will see diminished D2D performance and virtually no deduplication.  Sure, 512k may be a tad better, but 256k is still good.

 

For those running 11i v3 (11.31), it's somewhat of a mixed bag.  To guard against unpleasant surprises that may result from the quiet "chunking" noted above, the latest MA patch for DP has imposed an artificial block size limit of 256k for HP-UX.  That's good except that the stape driver in 11.31 can in fact handle larger block sizes like 512k.  Your options then if your media host is HP-UX 11i v3 are:

 

  1. Stay at a 256k block size with tape devices (D2D and physical).
  2. Remove the latest MA patch (PHSS_41963) and revert to the next-to-latest (PHSS_41460).
  3. Keep the latest MA patch, open a support call, and request the SSP for QCCR2A31441.

 

Options 2 and 3 will allow you to successfully use a 512k device block size with DP and HP-UX 11i v3.  From reading the lab case notes, it looks like the SSP provides updated BMA and MMA binaries that will not impose the 256k limit on HP-UX 11.31 where such restriction is not required.

 

Media Agent host is MS-WIndows

Here's where things may get a little cloudy.  If you experience an error in DP attempting to use a tape device configured with a 512k block size, the cause is most likely a configurable parameter called MaximumSGList which is used by the service for your Fiber Channel (FC) Host Bus Adapter (HBA).  A little sleuthing is required to track down the proper service to modify in the registry.  A quick trip to Device Manager will give us a start.  Expand SCSI and RAID controllers then locate the FC controller(s).  In my example, we have a pair of QLogic HBAs.

 

Dev_Mgr.jpg

 

 Check the properties of the HBA and look specifically for the Service property on the Details tab.  Here we see a value of ql2300.

 

HBA_Props.jpg

 

That tells me I need to dive into the registry at \HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\ql2300\Parameters\Device

 

Reg_Edit.jpg

 

Note that there is no current key named MaximumSGList.  In the absence of this key, the service will be operating at a default Scatter/Gather value.  On this particular host, I happen to know that the default value will support up to a 256k block size.  If we wanted to allow operation at 512k, it would be necessary to add a new DWORD "MaximumSGList" with a decimal value of 129.  The formula is (BlockSize/4096) + 1.

 

Mention is also made of MaximumSGList in DP's help facility.  For a direct hit, search on example of changing block size on windows media agent.

 

Help_Example.jpg

 

Note that a restart is necessary to implement the new value before testing.

 

Reg_Edit2.jpg

 

With our addition made and the reboot complete, I reconfigured one of the D2D VTL drives (LTO4 emulation) to use a block size of 512k.  The format of a virtual cartridge was successful at 512k and so was the subsequent backup of data to the cartridge.

 

Conclusions

It is possible to achieve a 512k block size with MS-Windows and HP-UX 11i v3.  I haven't addressed Linux as a Media Agent host primarily because it wasn't possible in the test environment to which I am so graciously allowed access.  It's also important to note that you shouldn't go crazy with MaximumSGList and arbitrarily set it to some really high value.  As noted in Microsoft Support Article ID 280793:

MaximumSGList is a value between 0 and 255. Use this adjustment sparingly as scatter/gather lists consume nonpaged memory.

My profound gratitude goes to a number of people who responded to my internal queries:  Christian, Paul, Andreas, Jaime, Joerg, Mike, Nikhil, Harold, and anyone I might have missed.  It is indeed a pleasure to work with such knowledgeable and helpful folks.

 

Have we missed anything?  Do you have experience that can augment or clarify what is documented here?  If so, please share with us using the comments section so that we may all benefit!

Labels: Data Protector
Comments
Jenni | ‎08-16-2011 09:14 AM

Mr T,

 

Great blog and thanks for all the screenshots! Convincing customers to change variables on their production servers (that are also MA's) via the registry can be challenging, especially if they decide to consult with the Storage Team first - your blog made this easy for me and the customer was also a lot more confident about what he needed to change and why.

 

jen

Mr_T | ‎08-22-2011 05:09 AM

Thank you for taking the time to comment, Jen.  I'm very pleased to know that the information was beneficial.  Please share the URL hp.com/go/dptips with your colleages and customers.  The more we work together, the more we all win!

sc_2111 | ‎11-10-2011 11:17 AM

how about BS of 1024 in a environment with windows MA ?

And also can you provide recommendation for segment size ?

thanks

 



Dominik Kowalski(anon) | ‎11-29-2011 12:33 PM

I tried to fight larger block size then 256kb on Linux Media Agent host on DP 6.11, maybe you can provide me some info about that? First try was setting ST_MAX_SG      256 do 512 value in "st" tape driver , but it didn't work...

Dominik Kowalski(anon) | ‎11-30-2011 07:58 AM

Hello there!

 

Thank you for a such a great article... I try to fight the same on Linux Media Agent hosts... and don't have luck, still stuck to 256KB on VLS system. I tried to look on st driver sources and give a try to set up:  #define ST_MAX_SG      256 to 512 in st_options.h and "max_sg_segs:Maximum number of scatter/gather segments to use (256) (int)" setting via insmod also, no luck either. Can you provide me some directions how to fight the issue? Maybe some documents somewhere?

Rasser(anon) | ‎02-06-2012 05:41 PM

Hi, I'm setting up a fully blown D2D4324 and dual site 3PAR SAN.

I was windering if it's possible to edit the blocksite via a script or something as otherwise I need to go through 60 odd drives? Have looked a bit at "omnidownload" but it doesn't seem any easier?

 

Great blog by the way!

Kota(anon) | ‎02-20-2012 02:41 PM

Hello,

 

I have a mixed set of tapes (mixed in terms of block-sizes) in my media-pool. Some are 64KB and some are 256KB. Is there any way to identify which tape is of which block size without affecting the data it contains?

 

Regards,

Kota :-)

Mr_T | ‎03-24-2012 03:35 AM

Hello all.  Sorry for my absence.  Let me do my best to answer recent queries.

 

Kota, I reckon you could try scanning each one first in a 64k drive then in a 256k drive to see which one doesn't throw an error.  Of course, if the tapes are scratch, why not just reformat all of them at the desired block size.  Another method is with TapeAnalyser.exe.  I did a bit of experimenting, and here's the command I honed to look at the first block of the first segment on a tape loaded in a drive:

 

E:\Program Files\OmniBack\bin\utilns>tapeanalyser.exe -device Tape0:0:1:0C -recdmp -range 1 1 1 1
======================================================

BLKSIZE = 65536

 You can use TapeAnalyser.exe -help for a listing of syntax and usage.

 

Rasser, I've faced a similar situation before.  Though the details were different, I found myself needing to update one parameter on dozens of drives.  I wound up writing a small shell script to omnidownload all of the device configs, another script using sed to make the updates, and then a third script to do the omniuploads.  I don't have them handy, or I'd list them here.  Thanks for your kind comments.

 

Dominik, I'm afraid I do not have a Linux box in the lab right now with D2D and physical tape, so I'm at a bit of a lost for testing.  You might pose your question on the Data Protector discussion forum:  http://h30499.www3.hp.com/t5/Data-Protector-Support-and-News/bd-p/itrc-251

 

sc_2111, larger block sizes are generally desired if you can achieve them.  The larger your block size, the fewer headers you will have for the same amount of data.  That's less backup app metadata for inline dedupe to sort through.

 

Thank you all for taking the time to comment!

BarryG(anon) | ‎04-12-2012 04:08 PM

Hi

 

I had the same problem recently with a DP6.2 Windows CM and Linux MA clients writing to a D2D.

 

The Windows clients were happy with the 512K block size but the Linux clients gave the error you describe in the Blog.

 

I got around it for any SAN attached Linux clients by adding the following entry in .omnirc on each client:

 

OB2LIMITBLKSIZE=0

 

Regards

 

Barry

Dan Crossman(anon) | ‎05-02-2012 08:44 PM

Hi -

 

I too was able to overcome the 256k block size limitation on Linux and DP6.2 with OB2LIMITBLKSIZE=0 in .omnirc. But on HP-UX 11.31, adding OB2LIMITBLKSIZE=0 to .omnirc only changed the nature of the error message

 

From this:

[Warning] From: MMA@bkup02.vcd.hp.com "HP:Ultrium 5-SCSI_3_bkup02"  Time: 05/01/12 11:38:44

[90:253]        Block Size setting of 512KB is too big. Maximum allowed block size for

        this device type is 256KB.

 

To this:

[Critical] From: MMA@bkup02.vcd.hp.com "HP:Ultrium 5-SCSI_1_bkup02"  Time: 05/01/12 11:40:11

[90:47]         Invalid device block size for HP-UX, supports only <= (256) KB => aborting 

[Major] From: MMA@bkup02.vcd.hp.com "HP:Ultrium 5-SCSI_1_bkup02"  Time: 05/01/12 11:40:11

        Initialization of medium failed.

 

(Incidentally, OB2LIMITBLKSIZE is not, by default, documented in .omnirc on DP6.2!)

 

I have doubts that the 256k MA block size limitation has been removed in DP6.2 for HP-UX 11.31.  

 

The only recourse at this point - to be able to run with 512k or greater block size and fully maximize performance and dedup on our D2D4324 - is to bring up a new Linux cell server to host the virtual tape libs for our hp-ux hosts.

 

Any ideas?

 

Leave a Comment

We encourage you to share your comments on this post. Comments are moderated and will be reviewed
and posted as promptly as possible during regular business hours

To ensure your comment is published, be sure to follow the community guidelines.

Be sure to enter a unique name. You can't reuse a name that's already in use.
Be sure to enter a unique email address. You can't reuse an email address that's already in use.
Type the characters you see in the picture above.Type the words you hear.
Search
About the Author
  • More than 30 years in Sales and Marketing in IT services business. Currently managing global campaigns for Datacenter Care.
  • I graduated in Software Engineering. Joined HP family five years ago, I deliver Insight Remote Support technical consulting for HP customers, in North America, Canada and Latin America. Assist setting up, installing and configuring the solution in customers' IT environments.
  • I am an identical twin. My brother’s name is Greg Tinker and we have been extremely fortunate working similar careers within HP, known to our HP colleagues and many of our customers as "The Tinkers". Our job is to be the technical lead on major business operational outages with millions of Dollars/Euros hanging in the balance. We both have a complete background in architectural, Infrastructure and application environments from both the proactive and reactive side of HP Enterprise Service (HP ES), and HP Enterprise Business (HP EB).
  • I am an identical twin. My brother’s name is Chris Tinker and we have been extremely fortunate working similar careers within HP, known to our HP colleagues and many of our customers as "The Tinkers". Our job is to be the technical lead on major business operational outages with millions of Dollars/Euros hanging in the balance. We both have a complete background in architectural, Infrastructure and application environments from both the proactive and reactive side of HP Enterprise Service (HP ES), and HP Enterprise Business (HP EB). We have always attended the same schools, studied the same material (big surprise, as we are identical twins), and have always worked as a close team and strive to demonstrate our teaming ability’s to others. We each have more than 11 years experience supporting mission-critical enterprise customers on a broad range of technologies. We’ve both won the HP MVP award multiple times as well as coauthored books, programs, and whitepapers in our spare time.
  • More than 25 years in the IT industry, managing ITSM, service development and delivery projects in Technology Services. Specialized in end2end support for ISV based business solutions. Certified ITIL and project management expert.
  • Eduardo Zepeda, WW TS Social Media Program Manager & Internal Communications for WW Technology Services Blogging on behalf of HP Technology Services (TS_Guest)
  • I have been with HP for 13 years, always in Services - first as a Services Channel Sales rep, then a Channel Services Segment Manager, and now, in WW Technology Services Marketing. These may be my formal job titles, but I'm really a Cheerleader for HP Services! I feel that HP has great services, exceptional Technical Experts and Delivery teams, and so many cool things are going on at HP Services. So, stay tuned...
  • I have 27 years of system, storage, and networking experience including detailed work with Data Protector (formerly Omniback II) for the past 14 years. My expertise includes StoreOnce deduplication technology, D2D appliances, performance tuning, complex remediation, and online backup integration with applications like Oracle and infrastructure like VMware. Traveling across the United States and Canada as a Sr. Technical Consultant, I deliver specialized consulting for a broad variety of HP customers.
  • MrCollaboration (aka Jim Evans) is an HP Global Services Alliance Manager. He has worked in the IT industry for more than 30 years, 22 of which were spent with Digital Equipment Corporation, Compaq and HP. He works with many third party vendors and partners to develop processes to facilitate excellent support and service for mutual customers. Jim is also HP’s representative to the Technical Support Alliance Network (TSANet).
  • I've been working in Customer Service for over 20 years. During my career I've provided support services for Languages, Programming Libraries and Operating Systems. During the last 10 years I've provide support for Linux and more recently VMware. My current role is as a Technical Account Manager working in the HP Custom Mission Critical Services Industry Standard Operating Systems team. I provide both reactive and proactive operating system support for proLiant servers and blades. Our services in the Custom teams are built on statement of work contracts for large HP customers who need a customized mission critical support offering.
  • I've been working in HP since 2007 like IT agent, developer, Web designer and then like Web Project Manager
  • I like to listen as much as I like to talk. Why? My 25+ years in the technology industry has taught me that the key to delivering value to customers is to understand what they value in the first place! I developed this passion for customers and consultative selling during my 12 years with Accenture, and I have continued to approach customers in a consultative way during my 12+ year tenure with HP. I also have a passion for HP given my knowledge of our Product and Service Portfolio and the differentiators we possess that position us as a leader in the areas our customers are telling us they want to go. Converged Infrastructure, Converged Cloud, Big Data – and the associated Service and Support implications – all such exciting technology trends where our success will hinge upon our ability to differentiate ourselves versus others in the areas that matter most to our customers. Right up my alley, and I am proud to be part of the great HP team where I know we have the best solutions in the industry!
  • Tom Clement has over 30 years experience in the areas of adult learning, secondary education, and leadership development. During this time Tom has been a consistent champion of “non-traditional” training delivery methods, including blended learning, virtual delivery (self paced and instructor led), the use of training games and simulations, and experiential learning. Tom has spent the past 25 years of his career at Hewlett Packard, focused most recently on HP’s global Virtualization, Cloud, and Converged Infrastructure customer training programs. Tom manages the strategic direction and overall performance of these training programs, ensuring these worldwide programs help HP’s customers capitalize on the business opportunities made available by IT advancements in each of these subject areas. Tom and his global teammates utilize best in class instructors, course content and supporting equipment infrastructure to deliver these training programs to HP’s customers. The team prides itself on providing the Virtualization, Cloud, and Converged Infrastructure content customers need when and where they need it, anywhere in the world. Tom is based in the Washington, DC suburbs and can be reached at tom.clement@hp.com.
Follow Us