ADSM-L

Re: continue a restartable session

2002-10-29 12:18:22
Subject: Re: continue a restartable session
From: Alexander Lazarevich <alazarev AT HERA.ITG.UIUC DOT EDU>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Tue, 29 Oct 2002 11:16:07 -0600
adsmserv 3.1 on aix 4.3.3 machine.

doing full restores from client (dsm), is not working, it hangs, so now
i've tried forcing an incremental restore on a filesystem that is only
partly restored (7/20GB). this is the command im using from the unix
client:

dsmc restore "/home/mac/tate/*" -replace=no -subdir=yes

so it starts it, loads a tape, takes the tape out, then it just sits there
and hangs again. i can't even restore deaper subdirectories, they also
hang. the whole damn server hangs. this is my current status:

adsm> q se

  Sess Comm.  Sess     Wait   Bytes   Bytes Sess  Platform Client Name
Number Method State    Time    Sent   Recvd Type
------ ------ ------ ------ ------- ------- ----- --------
--------------------
     2 Tcp/Ip Run      0 S  691.9 K     102 Admin AIX      ALAZAREV
 1,526 Tcp/Ip Run      0 S   11.6 K     926 Admin AIX      ALAZAREV
 1,560 Tcp/Ip RecvW  8.7 M    2.2 M     615 Node  AIX   HERA.ITG.UIUC.EDU

adsm> q rest

  Sess    Restore        Elapsed    Node Name                    Filespace
Number    State          Minutes                                 Name
------    -----------    -------    ---------------------
-----------
    -1    Restartable         11    HERA.ITG.UIUC.EDU       /home/mac/tate


The only thing I can think of is i deleted a tape from the library
yesterday. i deleted it because the server told me to. it said this:

ANR8302E I/O error on drive TAPEDRIVE1 (/dev/rmt1) (OP=HANDLELOSTVCR,
CC=0,KEY=03, ASC=31, ASCQ=00,
SENSE=70.00.03.00.00.00.00.48.00.00.00.00.31.00.FE.0A.36.2C.10.10.00.09.01.31.-
08.0D.A8.60.41.A0.11.00.00.33.9B.CD.00.33.36.00.06.33.3D.00.06.00.00.00.00.01.-
00.D4.2B.3B.00.00.00.00.00.00.00.00.00.00,
Description=An undetermined error has occurred).  Refer to Appendix B in
the 'Messages' manual for recommended action.
ANR8831W Because of media errors for volume 16644F, data should be removed
as soon as possible.

so i migrated the data off the tape, then deleted that vol, and checkout
the tape. the database should pick up this information right? the database
should know that i've taken that tape out and any data on it has been
moved, right?

anyway, does anyone have any clues. my group is missing 13GB of data, and
im gonna be hosed unless i can get it off these damn tapes...

thanks in advance,

alex

On Tue, 29 Oct 2002, Justin Bleistein wrote:

> Alex,
>
>       With the level of TSM your at they're have been many issues with
> process and thread hangs but there is away around it. Oh yes my fine
> feathered friend!!!. Just go to the client enter the dsmc utility and type
> in cancel restore and select the restore number to cancel. Or do a cancel
> restore -1 on the server side either way will get the job done. Then
> restart the restore as follows: "dsmc restore "/filesystem/*" -replace=no
> -subdir=yes -any_other_otions_you_specify This will ensure an incremental
> restore. It's failing due to the resource wait timing out on the threads.
> This is a problem with older versions. If it still hangs send me you actlog
> if you would be so kind and I'll take alook at it for you. But this always
> worked for me in the past.
>
> Thanks!.
>
> --Justin
>
>
>
>                       Alexander
>                       Lazarevich               To:       ADSM-L AT VM.MARIST 
> DOT EDU
>                       <alazarev AT HERA DOT IT        cc:
>                       G.UIUC.EDU>              Subject:  continue a 
> restartable session
>                       Sent by: "ADSM:
>                       Dist Stor
>                       Manager"
>                       <[email protected]
>                       .EDU>
>
>
>                       10/29/2002 09:33
>                       AM
>                       Please respond to
>                       "ADSM: Dist Stor
>                       Manager"
>
>
>
>
>
>
> we've got ADSM 3.1 on an AIX 4.3.3 machine, with a 3575 tape library.
>
> we had a major hardware failure, and i'm restoring the last filesystem
> right now. i'm trying to retore a filesystem on an aix system, let's
> say its called /home/dude. its a 20GB filesystem. 7GB has already been
> restored a few days ago, but the session died because of a missing tape
>  which we do not have anymore (data not available to server) (don't ask
> why we dont have the tape anymore). but now i must restore the rest of
> that filesystem. there was a restartable session for this restore, which i
> canceled yesterday.
>
> since the restartable seesion is canceled. im just trying to restart a
> brand new restore for that filesystem. i go to the aix client, type dsm,
> then choose to restore /home/dude, then it starts working, and at some
> point it asks me if i want to overwrite, i say NO, then it hangs. its been
> sitting there for 90 minutes??!?!?:
>
> 1,380 Tcp/Ip RecvW  1.5 H  826.4 K   5.2 K Node  AIX    HERA.ITG.UIUC.EDU
>
> also, when i do a q rest i get this:
>
>   Sess    Restore        Elapsed    Node Name              Filespace
> Number    State          Minutes                           Name
> ------    -----------    -------    ---------------------
> -----------
>     -1    Restartable         88    HERA.ITG.UIUC.EDU      /home/mac/dude
>
> but the sessions should show up as ACTIVE, not Restartable.
>
> so how do i finish restoring this filesystem? i looked in the
> administrator guide, page 204, and it does NOT say how to start a
> restartable session. it only says how to cancel it.
>
> any ideas? should i just start over? i hate to do that if im just gonna
> have to wait another 90 minutes for it to start restoring.
>
> thanks in advance,
>
> alex
> ---                                                        ---
>    Alex Lazarevich | Systems | Imaging Technology Group
>    alazarev AT itg.uiuc DOT edu | (217)244-1565 | www.itg.uiuc.edu
> ---                                                        ---
>

<Prev in Thread] Current Thread [Next in Thread>