Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Sign In with Google

Poll

No poll attached to this discussion.

In order to better support our growing community we've set up a new more powerful forum.


The new forum is at: http://community.covecube.com


The new forum is running IP.Board and will be our primary forum from now on.


This forum is being retired, but will remain online indefinitely in order to preserve its contents. This forum is now read only.


Thank you,

Suggestions: manual entry of missing Hard Drive Serial Numbers, WireFrame 3D model physical location

edited June 2011 in DrivePool
Hi Alex, great podcast on thehomeserver show.   http://homeservershow.com/the-home-server-show-140.html


It was great to hear directly from you (StableBit Employee# 0000000001).  The show was very informative and I enjoyed learning about your complete background and way of thinking.   Listening to you swayed me to try Stablebit DrivePool over DriveBender.   (and I'll be buying Scanner as well, I'm sure)

I have few minor suggestions I'd love to see in Scanner/DrivePool. (or both?)

1) The ability to manually add the serial numbers of my drives if it they be queried via Scanner/DrivePool.   When most of my drives are the same model number, and serial numbers don't show up, it's confusing to which drive I'm replacing or monitoring etc.

2) The ability to show a WireFrame 3D model of ALL hard drives in the shape they are physically for easy visual aid 
(like the DiskManagement Add-in did on V1)
http://www.tentaclesoftware.com/products/whs/diskmanagementThis would compliment #1 above.  (again manually adding serial number if it can't be queried)I realize I could just put stickers near each drive, but this is more fun!  (and looks killer when showing off the server)The main reason is quick troubleshooting though so you know for sure where a drive is physically located. (all GUI)Who has time for stickers right?

More related details/question:
As you said, SMART info can't always be read on certain non standard USB to SATA controllers/chip sets Could the same be true about not being able to read serial numbers via drives in my ESATA external enclosure?
 I have an HP EX495 running WHS 2011.   I have an external 5 bay SansDigital enclosure hooked up via ESATA cable.http://www.sansdigital.com/towerraid/tr5mb.html


If I recall, any time I used a Drive info type Add-In for WHS V1, the app couldn't read the serial numbers on every drive.   (including Scanner, SMART, among others)   My theory is that somehow the serial number is not able to be queried via ESATA to the SansDigital.  I never had time to investigate which drives were not showing serial numbers, but I'm pretty sure it was ones in the SansDigital.

I recently had the WHS V1 OS drive crash on my EX495.  Trying to recover failed so I said F'it and put WHS 2011 on it.   The USB install didnt work, so I read I can pre-install WHS 2011 on another machine and than transplant it.  So I hooked up a new spare drive to my regular desktop machine, installed WHS 2011, turned it off right before the first reboot mid-install.   I then put the drive in my EX495 and it booted up, finishing the install headless!  (was able to http://media/connect)  Success I'm up and running.

Now that I'm on WHS 2011 I'm excited to get Drive Pool capability back.   I was able to get all my data back from the hidden DE shares on the old DE drives.  (plus I have backups elsewhere)   Lucky for me I have a whole box of spare RMA'd 1.5 & 2TB drives thanks to a friends server business (he can't re-use) 
Question, can we add drives to the pool that have existing data on them?   It would save many many hours of copy/transfer time if we can.   My videos / ISO's are split between 4 drives right now, so hopefully I can combine them with ease.  If not, oh well I'm ready!   I think I'll wait for M3 so I have the minimal functionality I need.  (namely default folder redirection) 

One final low priority request:3) Maybe you can add a WHS Add-in so we can run Seagate SeaTools (or other OEM RMA scanners) via WHS?   This is probably a bit much to ask since it's not that hard to run it via Remote Desktop etc.   I'm thinking  maybe a simple link you can click on and the SeaTools app passed through the dashboard (as if it was remote access)  It seems the new server remote access has more options for something such as this maybe?

Ok well I didn't mean to ramble on and on here but thanks!   (looking forward to testing starting with M3)

~JazJon

Comments

  • edited June 2011 Covecube
    Thanks. It was fun doing it :)

    1. Good idea about the serial numbers, perhaps in Scanner 2.0 that's coming to WHS 2011. Probably not in the first release because there's enough work to do there already.

    2. That's a great add-in. I wonder if they'll do a WHS 2011 version. I don't want to duplicate functionality and rather develop unique features if they will. If not, perhaps post 2.0 release of the Scanner.

    Ah, the wonderful world of Windows I/O.

    Here's the short story with serial numbers. Simply stated, for SATA/PATA drives, the serial number is gotten by sending an IDENTIFY command to the drive and reading a response. Now if that worked for every drive then we would have serial numbers everywhere.

    Now you send commands to your drive all the time. Every time you read or write to a disk you are sending commands to it. So what's the big deal you may ask?

    Well... 99% of the commands that your drive services are read / write commands so those MUST work or else your drive would not do much. And the remaining 1% are all the other commands, including SMART. As you can imagine, implementing support for these is not a priority for driver writers / hardware makers.

    This situation is improving as people move to more modern operating systems, which have better APIs.

    When an application tries to read some data from your drive, the command is translated into a SCSI-like language in the Windows kernel, then the driver translates it back into ATA commands and send them down to the drive. This also changes depending on what architecture the driver is using SCSI-Port (legacy) StorPort ATAPort. It's a mess.

    With USB it's different. The I/O requests are translated into a SCSI-like language by the Windows USB storage driver (which is thankfully the same on all machines) and the bridge chip in the USB enclosure then takes those USB command codes and translates them to ATA commands to send to the drive.

    There are a lot of different layers involved and it may be different on each machine depending on the hardware / drivers installed.

    The problem is this:

    In order to get the serial number reliably, each piece of software and hardware that does all these translations needs to know how to translate that IDENTIFY command.

    Now Windows itself has many API calls to try and get at the serial number, it has WMI which keeps track of this stuff, and if that doesn't work you can send control codes down to the individual drivers.

    StableBit Scanner takes care of all this magic automatically. It doesn't work for all drives / enclosures, but it does its best. In particular, RAID systems tend to not support targeting individual drives in a RAID system.

    The bottom line is that if you don't see the serial number, then the Scanner tried WMI, it tried dealing with SCSIPort/StotPort/ATAPort drivers as best as it can and it didn't get anywhere.

    2.5. Adding drives to the pool with existing data?

    Adding a drive to the pool with existing data on them is supported. That data will NOT be made part of the pool.

    When you add a formatted drive to the pool, nothing is erased.

    3. Not sure how that would work with the distribution rights, etc... Perhaps a link could be shown to the OEM scanner when something goes wrong.
  • edited June 2011 Member
    Interesting, good to know stuff.

    Ok so I know I said I was going to wait for M3 to start testing Drive Pool but I dove in today with M2!  So far so good, having fun making a big pool again.

    I made a spreadsheet with the location of all drives and serial numbers
    So I have more related questions that are probably more in the Scanner catagory than drive pool, but I thought I'd add it here since we have all the details above already.

    I have the EX495 with 4 drives and a TR5M-B connected via ESATA cable with 5 drives.

    SeaTools can only run tests on 4 drives in the EX495.
    (both model and serial number are listed)

    All 5 drives in my external San Digital TR5M-B show up in SeaTools under: SCSI-FC-SAS 
    (model number is listed but serial numbers are not)
    SeaTools says test unavailable when I try any of the TR5M-B drives.

    If SeaTools can't even let me run the SMART test on drives in my TR5M-B, how will Scanner or Drive Pool ensure my drives are properly monitored for pending problems?

    I plan on buying the new AWIECO home server

    and updated newer TR5M-BP
    (better speed specs, possibly more compatible with SMART etc)

    Do you think the newer TR5M-BP might pass serial number/SMART data to SeaTools better? (and Scanner 2011)


    Drive Pool Questions...............

    EDIT:  I read the details technical overview and it answered a lot of the questions, but would still love to hear more.  http://blog.covecube.com/2011/03/stablebit-drivepool-technical-overview/

    1) Can I move drives around without confusing Drive Pool?
    Physically moving order drives around as well as moving from TR5M to EX495

    2) Changing Drive Letters.  (say I want E: to be Z: after the Drive Pool is already going)
    Does DP use Drive Letter to identify the pool or just the hidden directory name? 
    I'm also changing the label of all drives to DP1, DP2, DP3 etc to help keep track.  I'll start at the end of the alphabet etc of course.   Just curious if I can rename drives and move them around in a more logical "physical" order as I slowly add/change drives

    3) Will be be able to switch a Drive Pool Folder from Duplicated to non-Duplicated (and vice-versa) in future builds?   Hopefully the final Drive Pool and or Scanner will break down in pie chart pooled-regular vs pooled-duplicated usage etc too.

    4) I'm assuming performance is probably better on the drives inside the EX495 vs the drives inside the external TR5M enclosure.  Will/Can Drive Pool take this into account and optimize the DP to use the best performing drives first?

    Keep up the good work, it's because of you we can all enjoy the benefits of WHS 2011 finally.

  • edited June 2011 Resident Guru
    I'm not Alex, but I can answer three of your DrivePool questions:

    Q0) Can we add drives to the pool that have existing data on them?
    As Alex said, added drives keep their data but it is not automatically migrated (that feature is planned).

    DIY migration (i.e. move rather than copy) is possible, see http://forum.covecube.com/discussion/comment/245#Comment_245 (this is still valid as of the current build #1856).

    Q1&2) Can I move drives / change drive letters without confusing DrivePool?
    What about drive labels?

    A) Stop the StableBit DrivePool services, make the change, restart the services. Done.
    Labels can be changed even while the DrivePool is running.
  • edited June 2011 Covecube
    3) Will be be able to switch a Drive Pool Folder from Duplicated to non-Duplicated (and vice-versa) in future builds?   Hopefully the final Drive Pool and or Scanner will break down in pie chart pooled-regular vs pooled-duplicated usage etc too.

    4) I'm assuming performance is probably better on the drives inside the EX495 vs the drives inside the external TR5M enclosure.  Will/Can Drive Pool take this into account and optimize the DP to use the best performing drives first?

    Keep up the good work, it's because of you we can all enjoy the benefits of WHS 2011 finally.

    3. Yes M3 will have all that.

    4. Yes it can, but no it doesn't :) There are more important things to do now, like basic file balancing to begin with. M3 will have a file balancer. It will be as out of the way as possible (i.e. it will not run every 30 minutes). It will not optimize for bus speed, as you suggest, although we can do this in the future.

    Shane,
    Thanks ;)
Sign In or Register to comment.