block size vs. blocking factor

block size vs. blocking factor

Post by Mike Barnar » Fri, 24 Jan 1997 04:00:00




   (Alan Rollow - Dr. File System's Home for Wayward Inodes.) writes


writes:

>>After looking over the man pages for st and ufsdump in
>>Solaris, I am still confused  as to what the differendces
>>are between a block size for a tape and
>>the blocking factor for the tape. Any tape gurus answer
>>this quick one for me?

<snip well-written explanation>

Quote:>If there is a difference in the block size and the blocking factor,
>this is probably it.  One is the I/O size used by the application
>(ufsdump) and the other is the internal block size that the drive
>uses.

My 2 cents worth:

On some HP MPE machines I used to work on, "Blocking Factor" was
a number specifying the number of records to write to each block
on the tape.  Block size was simply the BF multiplied by the record
length.  For example, a BF of 20 with a record length of 500 gave a
block size of 10000.

Some further clarification on the efficiencies affected by BF
(again, at least in the context of the HP 3000's I used)

As a tape is written from the machine, a block's worth of data is
read from the disk file, then written to the tape.  Consequently,
smaller block sizes (which result from lower BF's) result in more
reads to transfer a file, and consequently much more time spent
on disk to tape I/O.

Also, each block on the tape is separated by physical space, so
more blocks on the tape (again, from a low BF) result in
more "unused" space - tape not occupied by data.  I've had tapes
run out of room from use of a too-low BF (and take
much too long to write), that were able to hold the file(s) when
the BF was increased (more data per block = less tape
used for physical space between blocks -> the file fit on the tape)

Mike

My opinions are mine and not necessarily those of my employer, who
has no opinions except when it comes to those of its employees...

 
 
 

1. SCO tar error on block size (not block factor)

Someone recommended:
tar cvfb - ./config 32768 | compress | dd of=/dev/rmt/0m bs=32k  
tar: invalid blocksize. (Max 20)
0+1 records in
0+1 records out

Seeing his/her error i supposed that what they meant was the actual
block
size in bytes to match the 'dd' record size.  But
tar cvfB -  ./config 32768  | compress | dd of=/dev/rmt/0m bs=32k
tar: B: unknown option
Usage: tar -{txruc}[0-9vfbkelmnopwAFLTP] [tapefile] [blocksize]
[tapesize] files...
        Key     Device            Block   Size(K)    Tape
        0       /dev/rfd048ds9    18      360        No    
        1       /dev/rfd148ds9    18      360        No    
        2       /dev/rfd096ds15   10      1200       No    
        3       /dev/rfd196ds15   10      1200       No    
        4       /dev/rfd0135ds9   18      720        No    
        5       /dev/rfd1135ds9   18      720        No    
        6       /dev/rfd0135ds18  18      1440       No    
        7       /dev/rfd1135ds18  18      1440       No    
        8       /dev/rct0         20      0          Yes    
        9       /dev/rctmini      20      0          Yes    
        10      /dev/rdsk/fp03d   18      720        No    
        11      /dev/rdsk/fp03h   18      1440       No    
        12      /dev/rdsk/fp03v21 10      20330      No  

Is there a way to get tar to match the 'dd' record size please?
--

have
access to a news server; thanks!
Disclaimer: opinions expressed my own and not representative of my
employers

2. Upgrade cards and Linux PPC

3. Non blocking socket blocks; says 'read would block' ?

4. linking two or more machines?

5. EXT2: Defualt block-size & super-block pos.

6. --Still having PPP problems. HELP!!!

7. Finding the block size of a file on tape - block.c (0/1)

8. X hangs whole System

9. Does TRU-64 support devices with block sizes > 512 bytes/block ?

10. Disk Block size vs Bytes

11. logical vs physical block size

12. Logical block sizes (newfs/mkfs vs mount)

13. DAT blocking factor & tapetool