Change the Block Size

[Rogelio Cabrera]

I have 4 disks in my serer hp9000 with hp-ux 11.0 and After Executing the tool mstm I saw that it has 512 how block size.
My Question is: How Change this size??

I need delete the lvol at my disk and all the information???
A procedure exists to move the size without delete the info???
Thanks and Excuse me for my bad english.
Greetings.

[Steven E. Protter]

Shalom,

Block size is set at volume group creation.

You can probably back up the data, vgexport and vgimport.

A. Clay Stephenson has done some performance studies based on block size with oracle and found it has little impact on performance.

You English is fine. Better than my Hebrew for certain.

SEP

[Patrick Wallek]

The block size here cannot be changed easily.

As far as I know all disks use 512 block size (or sectory size) UNLESS you are using a VA disk array which uses 520 byte sector sizes.

I don't think there is much point in attempting to change this.

[Bill Hassell]

mstm is reporting the classic Unix block size of 512 bytes but this has nothing to do with the size of data blocks used by applications. In the really old days, disks were very slow and very dumb. They had little or no local RAM for buffering so it was important to have the filesystem control blocks of data regardless of what the applications wanted. The HFS filesystem has the concept of a block and a fragment, but all of this disappears in VxFS filesystems where data blocks are dynamic. And the blocked disk driver also adjusts the actual block size by concatenating smaller records together before transferring them to/from disk.

[Rogelio Cabrera]

I have Progress Data BAse Ver. 9.1d and i read a manual where they comment that it is necessary to take care of the values of block size at FS, but not of disks.

It already reviews the values of FS and are OK, They have 8192 COMO bLOCK SIZE.

THANKS TO ALL AND SALUDOS FROM MEXICO.