[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference ssag::ask_ssag

Title:Ask the Storage Architecture Group
Notice:Check out our web page at http://www-starch.shr.dec.com
Moderator:SSAG::TERZAN
Created:Wed Oct 15 1986
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:6756
Total number of notes:25276

6352.0. "Free bloks difference..." by SAFFO::BERETTA () Tue Jan 28 1997 08:43

	Why is there a discrepancy between the total blocks utilization
	when I do the following two comands?
	In some cases (on other disks like stripeset ) the different consist
	of few milions of blocks.

	Any suggestion will be appreciated.

	$ DIR/SIZE/GRAND_TOTAL  $1$dia3:[*...]
	  
        Grand total of 223 directories, 3685 files, 2574452 blocks.

	$ SHOW DEV/FUL $1$dia3
                          
Disk $1$DIA3: (MLNGPX), device type RF36, is online, mounted, file-oriented
    device, shareable, available to cluster, error logging is enabled.

    Error count                    0    Operations completed              18407
    Owner process                 ""    Owner UIC                      [SYSTEM]
    Owner process ID        00000000    Dev Prot            S:RWPL,O:RWPL,G:R,W
    Reference count                3    Default buffer size                 512
    Total blocks             3125408    Sectors per track                    75
    Total cylinders             2605    Tracks per cylinder                  16
    Host name               "MLNGPX"    Host type, avail      VAX 4000-200, yes
    Alternate host name     "SOSGPX"    Alt. type, avail      VAX 4000-200, yes
    Allocation class               1

    Volume label             "USERS"    Relative volume number                0
    Cluster size                   3    Transaction count                     3
    Free blocks               493173    Maximum files allowed            390676
    Extend quantity                5    Mount count                           3
    Mount status              System    Cache name          "_$1$DIA2:XQPCACHE"
    Extent cache size             64    Maximum blocks in extent cache    49317
    File ID cache size            64    Blocks currently in extent cache      0
    Quota cache size               0    Maximum buffers in FCP cache        377
    Volume owner UIC        [SYSTEM]    Vol Prot    S:RWCD,O:RWCD,G:RWCD,W:RWCD

Volume Status:  subject to mount verification, file high-water marking,
write-
      back caching enabled.
  Volume is also mounted on SOSGPX, MLNGPX.       

	
	In cases of few millions of blocks of difference, there are also 
	performances problem to acces to that specific disks and the 
	workaround was to force a rebuild by SET VOLU/REBUILD.
	

	Thanks			
				Claudio
T.RTitleUserPersonal
Name		DateLines
6352.1FILES vs. BLOCKSSUBSYS::VIDIOT::PATENAUDEAsk your boss for ARRAY's...Tue Jan 28 1997 09:358
    
    Ahh, let me guess,
    
    probably because the DIR command is only telling you the total size of
    all the FILES on your disk and the show dev is showing the total BLOCKS
    for the disk (which by the way is correct at 3125408)
    
    roger.
6352.2AMCFAC::RABAHYdtn 471-5160, outside 1-810-347-5160Tue Jan 28 1997 10:4513
re .0:

First and foremost, you must specify /SIZE=ALLOCATION, otherwise you are only
seeing blocks in use.  Look at the results from using;

	$ COPY NL: X.X/ALLOCATION=100

That'll help a lot but you'll still be missing some blocks -- specifically,
those belonging to the files in $1$dia3:[000000] and those, if any, not in any
directory at all (i.e. lost -- use the ANALYZE/DISK utility to look for and
recover them).  Also, watch out for files with two or more directory entries --
they are only counted once via SHOW DEVICE/FULL but repeatedly via the DIRECTORY
command.
6352.3Asked and answered 1,000 times in VMSNOTES!!!COOKIE::FROEHLINLet's RAID the Internet!Tue Jan 28 1997 13:395
    Clearly a hardware problem. You have to open the drive and cleanout
    the free block count by hand. Make sure you touch all platterns. After
    this procedure your freeblock count problem will be clearly gone. ;-)
    
    Guenther
6352.4SUBSYS::VIDIOT::PATENAUDEAsk your boss for ARRAY's...Tue Jan 28 1997 14:483
    
    ;^)
6352.5SUBSYS::BROWNSCSI and DSSI advice given cheerfullyWed Jan 29 1997 09:3113
    Come on guys, answer the question in clear English, without sarcasm or
    obscure idioms.  The question has undoubtedly been answered in 
    VAXAXP::VMSNOTES, but finding it amid the 12-volume archive is not easy.
    
    
    The simple answer to .0 is that you're comparing the wrong numbers.
    Dir/size/grand_total shows about 2.6 million blocks in use.
    Show device/full says the device has 3.1 million blocks, of which
    500,000 are free.  So ,there are about 3.1 - 0.5 = 2.6 million in use.
    
    There is a small difference between "dir/size" and "show device/full",
    about 50,000 blocks.  If you use "dir/size=all", as recommended in 
    a previous reply, the difference will disappear.
6352.6COOKIE::FROEHLINLet's RAID the Internet!Wed Jan 29 1997 11:1017
    Come on guys, this was one of my light jokes. ;-)
    
    .0 indicates that this is a cluster. In which case the free block count 
    in the "SHOW DEVICE" can be wrong all by itself. If that's the case
    a clustewide re-MOUNT of the volume changes the free block count to a 
    more truely value. Discussions on archived VMSNOTES conferences show
    all the details and aspects of this issue. A "Notes> DIR/TITLE="FREE BLO"
    (be careful with English spelling) showed:
    
    	VMSNOTES_V12	 917
    	VMSNOTES_V11	1929
    	VMSNOTES_V10	 627
    	VMSNOTES_V9	 940 2405 2919 3224 4036
    
    Took me about 5 minutes to find all these valuable discussions.
    
    Guenther