gconcat file system damage

[Alban Hertroys]

On 11 Sep 2009, at 5:45, Andrew Snow wrote:

> Peter C. Lai wrote:
>> What is the status of growfs(8) then?
>
> As far as I can tell, it doesn't work reliably with UFS2 partitions,  
> and it doesn't work at all with large partitions.
>
> People who do try to use it, can end up with corrupted  
> filesystems... and the code is currently unmaintained.

Aha, I forgot about that step, but I did indeed use growfs to expand  
it. Is there any way to repair the corruption without erasing my data,  
or do I need to buy an external disk to backup to first (which I need  
for backups anyway)? If so, what's the pattern of the corruption? Is  
it likely to only start after a certain size?

Lastly, having a tool in the base system that causes these kinds of  
issues doesn't seem like a very good idea. growfs does issue a warning  
that made me read the bugs section in the man page. That section  
points to an ffsinfo command that I ran on the filesystem in question,  
but it appears to be alright:

===== START CYLINDER SUMMARY =====
# 0 at 28202000: 0. csum in fscs
ndir   int32_t 0x00000001
nbfree int32_t 0x00002474
nifree int32_t 0x00005bfd
nffree int32_t 0x0000000d
===== END CYLINDER SUMMARY =====

The filesystem is 740G, which seems to fit in the 'critical data  
structure'; 1480 < 0x2474...

Or is this a different (and undocumented) issue?

Alban Hertroys

--
Screwing up is the best way to attach something to the ceiling.


!DSPAM:74,4aaa10f412071914714074!