Subject: | |
From: | |
Reply To: | |
Date: | Mon, 29 Jul 1996 13:31:34 -0700 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Paul writes,
> Do you mean 20 extent blocks or 20 extents? There is a limitation on
extents...based on memory of old conversations and pre-non-disclosure
knowledge.
> Wait. 768 pages is indeed .75 MB. Except the very first extent is 769
> pages (one more page for the header), all other extents are .75MB.
768 * 4096 = 3145728 (all in decimal)
(see attached example)
BTW, that's a lot more than .75 MB ... it's 3 MB.
> Label table is very picky about this. Everytime the label table expands,
> it want EXACTLY .75MB.
Here's an example...I have a label table with 1 extent, and about 600
unused entries...so I created a bunch of files and looked again...2 extents,
and the second one is 3 MB.
I used no programs to verify this...just debug and watching the #extents
in the Label Table's extent entry block.
> It is indeed possible to see label table extent size differ than .75MB.
> It is an indication some one or program had played with the label table.
well, at least one machine has a freshly INSTALLed system, with no
utility of any kind run on it (LPS, Bradmark, or HP!)...and 3 MB
label table extents.
> Label table is just a file. If someone is reading or writing it beyond
> its current EOF, it will cause an extent fault and produce a different
> sized extent! One quick way to do it without even knowing is to look
> at the label table beyond EOF with DEBUG! Don't try this at home though,
> it waste disk space and slow down bootup time!
Yep.
------------------------------cut here-----------------------------
De-Frag/X B.xxx (run on 7/29/96 12:28:40.1)
Copyright (c) 1993 Lund Performance Solutions
...
De-Frag/X> de ldev all 1
-------------------------------------------------
Ldev 1, entry# 1 --> $11.$600 = $LABEL_TABLE
File: $LABEL_TABLE (2 extents)
Ldev Disk Page# # Pages File Page #
---- ---------- ------- -----------
1 133 769 0
1 26,482 768 769
# extents in file: 2
# pages in file: 1,537 pages; 6.0 MBs
% fragmented: 0.1
...
from Debug/iX:
fv 11.600 "t_file_label_entry,true"
RECORD
VERSION : 10
LT_ENTRY_TYPE : LT_FILE_LABEL
PHIL : 0
FCTS0 : b0
FCTS1 :
DAY : 88
HOUR : 2
MINUTE : 4
SECOND : a
TENTHS : 1
LAB_ENTRY_OFFSET : 600
EXTENT_BLK_OFFSET : 800
INTERVAL_TIMER : 1c6b6dd
II_FLAB_EXT_OFFSET : 0
...!!!
fv 11.800 "t_extent_block_entry,true"
RECORD
VERSION : 10
LT_ENTRY_TYPE : LT_EXTENT_BLOCK
NUM_EXTENTS_IN_BLK : 2
FCTS0 : b0
FCTS1 :
DAY : 88
HOUR : 2
MINUTE : 4
SECOND : a
TENTHS : 1
PARENT_LABEL_OFFSET : 600
PREV_EXTENT_BLK_OFFSET : 0
EXTENT_BLOCK :
[ 1 ]:
EXTENT_SECTOR_ADDR : 850
SECTORS_IN_EXTENT : 3010
VOL_SET_INDEX : 1
FILE_SECTOR_OFFSET : 0
[ 2 ]:
EXTENT_SECTOR_ADDR : 67720
SECTORS_IN_EXTENT : 3000
VOL_SET_INDEX : 1
FILE_SECTOR_OFFSET : 3010
[ 3 ]:
EXTENT_SECTOR_ADDR : 0
SECTORS_IN_EXTENT : 0
VOL_SET_INDEX : 0
FILE_SECTOR_OFFSET : 0
...
= 3010*#256, #
#3149824
--
Stan Sieler [log in to unmask]
http://www.allegro.com/sieler.html
|
|
|