Hi Stuart.

On 12/05/2009, Stuart Henderson <[email protected]> wrote:
> I use a bunch of CF based systems, and mount most of them read-write.
> I've done so for years, and haven't had any problems as a result.
>
> I've had cards fail, but mostly after just a month or two, and other
> identical cards bought at the same time in the same conditions (similar
> number of write/erase cycles) have been totally ok, so I'm putting that
> down to individual bad cards.

I will stop stressing then.

>> Hey I solved my first problem (I think).
>> I was going to ask how to populate (-P) the mfs mounts without having
>> to duplicate the original directories on my CF (/populate).
>> I think I could populate them straight from the original directories. :]

I'd like a recount on that. Obviously it will not work. Haha.

>> Anyway, are there other ways to populate mfs mounts with system files?
>> Is there any way to mount only the specific files I need?
>> For instance if I never use a specific log file is there any way to
>> not mount it?
>
> No.
>
> You could adjust the path of some file in syslog.conf and
> newsyslog.conf though. There is also the option of using a circular
> memory-buffer for logs, see syslogd(8) -s, syslogc(8), syslog.conf(5),
> but you probably need to write to some other things in /var (e.g.
> /var/run, /var/tmp; they could be symlinks though).

That's the kind of road I was thinking of.

> At least if a file is never used, it's not going to take a lot of
> space on the ram disk :-)

True.
It bugs me to duplicate every file in /var.
Especially things like /var/www/icons ...
It really bugs me to duplicate /dev and /var on the CF to be able to
copy them to RAM.
Triplicate.

I think some fancy script is what I need.
MAKEDEV to my RAM /dev and changing the paths to the log files.
I know I won't sleep well till it's done. :]

Thanks for the clarification.

Best wishes.
_______________________________________________
Openbsd-newbies mailing list
[email protected]
http://mailman.theapt.org/listinfo/openbsd-newbies

Reply via email to