[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#381982: XFS and hosts problems (Re: Bug#381982: Installation Report on AMD64 Laptop called deen)



Hallo,

On 8/22/06, Nathan Scott <nathans@debian.org> wrote:
On Mon, Aug 21, 2006 at 10:57:46PM +0200, Oleg Verych wrote:
...
> in prev. email, result is nothing from xfs_check and xfs_repair.

repair and check wont be reporting a problem because the size is
valid... it looks like somethings has done a truncate(2) at a large
offset (at least to the xfs tools).

what do you mean "size is valid", ls, du, dd ?
,-
|deen:/var/log# ls -hl lastlog faillog
|-rw-r--r-- 1 root root 605M 2006-08-22 12:57 faillog
|-rw-rw-r-- 1 root utmp 5.4G 2006-08-22 12:57 lastlog
|deen:/var/log# du -h lastlog faillog
|8.0K    lastlog
|5.0K    faillog
|deen:/var/log# dd <faillog>/tmp/faillog-copy
|dd: writing to `standard output': No space left on device
|511177+0 records in
|511176+0 records out
|261722112 bytes (262 MB) copied, 2.67818 seconds, 97.7 MB/s
|deen:/var/log#
`-
Well, dd doesn't do IO on hdd and xfs_bmap says there are huge holes.
How come ?

Pattern inside files is "0x00", before any records. Maybe some there are 64bit
overflow bugs in "login" ?

,-
|deen:/var/log# xfs_bmap -vpl /var/log/faillog
|/var/log/faillog:
| EXT: FILE-OFFSET         BLOCK-RANGE      AG AG-OFFSET          TOTAL
|   0: [0..5]:             7019688..7019693  7 (5366..5371)           6
|   1: [6..7]:             7052964..7052965  7 (38642..38643)         2
|   2: [8..1238167]:       hole                                 1238160
|   3: [1238168..1238169]: 7050970..7050971  7 (36648..36649)         2
|deen:/var/log#
`-
what does "xfs_bmap -vp" on these files say?  was there an unclean
shutdown (and hence journal recovery) between the install and when
this problem was observed?

I've noticed something wrong when i've installed new debian-etch (with
non working
mkfs.xfs, you know this). Having new setup i also (loong ago) wanted new
partitioning on hdd. Copy (didn't want to waste time) of new setup was more than
1.2Gi, while i've installed minmial one ~200Mi (i have list of apps i
need). Thus i've
discovered those large *logs. So "wasting" time was inevitable, i've
tried gtk install
on new "for development" 1.5Gi partition -- same result with *logs.

All that files i've seen just after reboot to new installed system
(first and second
one worked for two days afterwards, without any reboots or errors). First (and
others) mounting is always clean.

--
Nathan


Thanks.

--
-o--=O`C
 #oo'L O
<___=E M



Reply to: