?
Solved

how to determine if a hard disk conform to SAS V2 ie that keep an error count of its bad sectors

Posted on 2011-04-19
13
Medium Priority
?
439 Views
Last Modified: 2012-05-11

is there a website that list out all the various makes / models of disks / HDDs
that tells us if a hard disk conforms to this standard?  Or how do I determine
if my SAN's HDD conforms to this standard?

I was told by a storage specialist that it's not a SAN feature but a feature of
hard disk drive that keeps track how many failed attempts it has tried to
read a sector unsuccessfully.
0
Comment
Question by:sunhux
  • 5
  • 4
  • 4
13 Comments
 
LVL 47

Accepted Solution

by:
David earned 800 total points
ID: 35424716
Your specialist is correct.   The retry settings are a programmable parameter in the disk drive.  Specifically, this is configured via a mode page editor.    Look at this online manual for a product and it shows you some of the settings on mode page 1 & 2.

http://www.santools.com/smart/unix/manual/example_mode_page_dump_sas.htm

The HDD manufacturer chooses whatever level of compliance they desire, and the http://www.t10.org, t11, and t13 sites publish the specs.

0
 

Author Comment

by:sunhux
ID: 35425554

If I provide the part number of my HP's EVA (Enterprise Virtual Array) SAN
fibre disk, is there anyway to tell if it complies to which level?
0
 
LVL 47

Assisted Solution

by:David
David earned 800 total points
ID: 35426002
Well, now I understand.  You probably have a "FATA" disk, which is actually a SATA drive behind a FC->SATA bridge chip.    That would explain why you can't access and/or adjust the timing.  SATA disks have no programmable equivalent.  

Now is it compliant?  Yes, actually it is, because if you look at the vendor-inquiry page, there are a few fields that report the ANSI compliance level and specifications the disk supports.  

0
VIDEO: THE CONCERTO CLOUD FOR HEALTHCARE

Modern healthcare requires a modern cloud. View this brief video to understand how the Concerto Cloud for Healthcare can help your organization.

 
LVL 47

Assisted Solution

by:David
David earned 800 total points
ID: 35426067
.. nope, that sort of stuff isn't published, except under NDA, (which I have, but I don't have the product manual and it would take quite some time to get it).
W/o the manual, I could certainly determine this in a few minutes running some code that enumerates all the mode, log, and SCSI EVPD pages, but I certainly can't (and won't) give you that code :)

What is the real problem?  Surely it isn't an argument about compliance?  
0
 
LVL 56

Expert Comment

by:andyalder
ID: 35434259
What a strange question, EVAs don't currently use SAS disks, although they're likely to in the future.
0
 

Author Comment

by:sunhux
ID: 35437958

Well, last month one of our EVA4400 fibre disk had a bad sector (according to HP & there was
no alert & the EVA failed to copy out data to another disk & caused the entire EVA to fail & the
entire DB became inaccessible.  

IBM just gave us a presentation & told us that the ability to detect bad sector depends on the
feature of the HDD & not the SAN.  So if the type of HDD used in the SAN complies to certain
standard (I heard it's SAS 2 if I did not hear wrongly), we would have been able to detect it
in advance.  What's more, IBM was presenting  "predictive"-type servers which will alert us
prior to actual hardware fault surfacing, so I'm wondering what type of HDD I currently have
in my EVA4400 & I can pre-empt future such bad-sectored HDD fault.  HP applied a firmware
patch later to the EVA4400 but I'm not sure if my other EVAs would face the same issue
0
 
LVL 47

Assisted Solution

by:David
David earned 800 total points
ID: 35438006
Bad-block detection on SAS / SAS-2 disks can be done via several techniques.  It is just a matter of sending the correct CDB to the device.  I won't bore you with the specific commands, as the ANSI spec can be looked up, but here are some big picture things
1.  You can query a disk at any time to report the specific locations of factory & grown defects (i.e, KNOWN blocks that have been remapped)
2.  You can issue a verify command that reads a range of blocks, in order to detect blocks that are unreadable at this moment.  This is a preferred method as it doesn't chew up bandwidth on backplane, and is done mostly in hdd firmware.  You could also just issue a read, but that requires significant overhead as this involves a data transfer.
3.  If disk supports BGMS, then the disk can search / repair bad blocks in background.  Most likely this is what IBM is talking about.  The feature was introduced, I believe, in the Seagate 10K model 5s and up.    Turning it on is no big deal, just need to do it with a mode page editor, and set the rescan interval.  the trick is getting the info and then doing something about it.  Since the disks are not JBOD and behind your EVA controller then you have to run code that can pass-through that controller.    One can get a list of bad blocks that has not been "discovered" by the RAID controller in 2-3 seconds, and program the disk to automatically retry during idle time to keep trying to re-read blocks internally to recover them.  

There are also a few dozen non-volatile counters one can access via LOG SENSE command. There are a half-dozen defined in ANSI spec relating to types of unrecovered and recovered read errors.  This is different then detecting bad blocks.

Also the SAS/SAS-2  spec lets you send SMF commands to get counters measured at the expanders, as well as initiators that report things like resets, & disparity errors.  This is all part of the ANSI spec.   So really lots of counters that have to do with "errors" one could access with the right code.  
0
 
LVL 56

Assisted Solution

by:andyalder
andyalder earned 600 total points
ID: 35439184
A bad block on a disk shouldn't even fail the LUN let alone the entire EVA. I suspect you were subject to the dreaded EVA4400 firmware bug, you can take "While we have had no negative reports from customers on the operation of XCS 09521000, our lab teams believe there is a critical improvement needed to the functionality added in support of the new Cache DIMM component." with a pinch of salt, customers don't report on the operation of firmware, they report that their freaking SAN crashed. It is essential to make sure you're not running that old version of XCS on any of your EVAs.


http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?objectID=c01850026&lang=en&cc=us&taskId=101&prodSeriesId=3664763&prodTypeId=12169 (now superced, but documents the unstable XCS version)
0
 

Author Comment

by:sunhux
ID: 35441482

Thanks very much gents.

Andy, my management now thinks that I should constantly scan HP ITRC for
possible future disruptive firmware bugs & then perform the upgrade.

Is it just a simple matter for end-user IT like me to do the EVA firmware upgrades
from HP StorageWorks Command View (I suppose this must be the Windows
box which I can group/ungroup disks & configure the EVA disk pools)?

Or such firmware upgrade is best done by HP considering that I have to check
the compatibility matrix (which a storage person in HP told me was only made
available to the public last year).  

I was told by the HP storage person that the compatibility matrix checks involve
checking if my fibre switches' firmware needs to be upgraded as well & whether
the FC drivers on my servers (Linux, Windows & HP-UX) needs to be upgraded
too prior to the eventual firmware upgrade on the EVA4400 (I suppose this means
upgrading the EVA4400's SAN controller firmware).
0
 
LVL 56

Assisted Solution

by:andyalder
andyalder earned 600 total points
ID: 35441895
The upgrade from XCS 09522000 was so serious that HP provided it as a free warranty service, not sure if that's still the case as it was a while ago. It's relatively easy anyway, especially if you use Command View EVA rather than the inbuilt management.

You don't actually have to keep monitoring for firmware and driver upgrades, you can subscribe your kit list and you'll automatically get emailed when anything comes out although 3/4 of it won't be relevant.

www.hp.com/storage/spock (free but you hve to sign up) has the compatibility matrix, and yes that used to be restricted to business partners, the info's always been available by drilling down to each bit of kit on the public drivers download site though.

Be sure to read the release notes if you upgrade the switches, it's not a case of banging on the latest firmware, it's a step by step upgrade so you have to upgrade it several times with little revision steps (at least with Brocade).
0
 

Author Comment

by:sunhux
ID: 35452517

Just 3 last questions :

1) does www.hp.com/storage/spock alert us via email as well when a new
    EVA firmware is released?  If not, does ITRC does this?

2) When I login to ITRC, I have quite some difficulty navigating to a place or
    link to sign up for email notification for EVA firmware releases.  If you can
    provide me step by step of what I need to click to get there, appreciate it.

3) I heard that firmware upgrade for HP's XP (OEM'ed from Hitachi?) SAN is
    not that simple for customers to undertake themselves, is that so?  So
    for EVA, other than the upgrade from XCS 09522000, firmware upgrades
    for EVAs are generally simple enough for customers to undertake on
    their own?  If the customer created a problem during an EVA firmware
    upgrade undertaken on their own, would HP support? (assuming the
    customer only sign up EVA hardware support with HP & no EVA software
    support)

0
 
LVL 56

Assisted Solution

by:andyalder
andyalder earned 600 total points
ID: 35452576
HP's web is confusing isn't it.

https://h30046.www3.hp.com/subchoice/subhub.aspx , enter email address and check the drivers and support alert checklist, click product in column A and it expands into column B, click subitem in column B and it goes into column C which is your product list. If you're already signed up then click "Returning users, click here to sign in" and there's a link on RHS to go to the subscription list.

Ignore the blue square that says signup: ITRC, that's for HP-UX.

I don't know XP range, yes, the hardware is OEM'ed from HDS but it runs DEC/Compaq/HP code. I can't answer the rest, you'll have to ask your reseller or HP.
0
 

Author Closing Comment

by:sunhux
ID: 35460009
ok
0

Featured Post

NEW Veeam Backup for Microsoft Office 365 1.5

With Office 365, it’s your data and your responsibility to protect it. NEW Veeam Backup for Microsoft Office 365 eliminates the risk of losing access to your Office 365 data.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Concerto Cloud Services, a provider of fully managed private, public and hybrid cloud solutions, announced today it was named to the 20 Coolest Cloud Infrastructure Vendors Of The 2017 Cloud  (http://www.concertocloud.com/about/in-the-news/2017/02/0…
Each year, investment in cloud platforms grows more than 20% (https://www.immun.io/hubfs/Immunio_2016/Content/Marketing/Cloud-Security-Report-2016.pdf?submissionGuid=a8d80a00-6fee-4b85-81db-a4e28f681762) as an increasing number of companies begin to…
This tutorial will show how to configure a new Backup Exec 2012 server and move an existing database to that server with the use of the BEUtility. Install Backup Exec 2012 on the new server and apply all of the latest hotfixes and service packs. The…
This tutorial will walk an individual through the process of installing of Data Protection Manager on a server running Windows Server 2012 R2, including the prerequisites. Microsoft .Net 3.5 is required. To install this feature, go to Server Manager…
Suggested Courses

862 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question