growfs filesystem size limits ?

[Arne Wörner]

--- Henry Vogt <hv at tuebingen.mpg.de> wrote:
> Am 18.05.2007 um 11:21 schrieb Arne Wörner:
> Now it worked! Many Thanks, the fs is now 6.6 TB (see below).
> The output is still somewhat strange, superblock numbers changing   
> from positive to negative and back.. but everything else looks ok.
>
Ohoh... I am afraid, that it damaged some data blocks now... :) Due to wrong
offset...

Although it looks like cylno maybe 32bit...
In line 262-263 it says:
      j = sprintf(tmpbuf, " %d%s",
          (int)fsbtodb(&sblock, cgsblock(&sblock, cylno)),
There we can see, that it casts the 64bit value to 32bit...

Maybe u r right and it is just the output...
It should be "%jd"... :-)

Efficiency increased from 20% to 30%... ;-) <--- taken from "1984 (Orwell)"...

> < static void   get_dev_size(int, u_int64_t *);
> - ---
>  > static void   get_dev_size(int, int *);
> 1927c1927
> < get_dev_size(int fd, u_int64_t *size)
> - ---
>  > get_dev_size(int fd, int *size)
>
These r important, too... :-)

> 1980c1980
> <     u_int64_t p_size;
> - ---
>  >     u_int32_t p_size;
>
There is somewhere an "unsigned int size;" or so... It should be "u_int64_t",
too...

> # df -h shows and it looks like everything is still there..
> 
> /dev/da1p1     6.6T    3.5T    2.5T    58%    /mnt
>
Have u done a "fsck /dev/da1p1", too?

-Arne


       
____________________________________________________________________________________Pinpoint customers who are looking for what you sell. 
http://searchmarketing.yahoo.com/