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

Re: cvs no space left on device




> This is not really a debian question, but I'll ask anyway. I ran out
> of disk space while retriving files with cvs. Since then I've freed up
> 2.5 Gigs worth of space, but cvs still says no space left on device. I
> have only one e2fs partition mounted as root, and it has only about
> 20% of it's inodes used up.
> 
> The entire response from cvs will be like:
> 
> can't create temporary directory /tmp/cvs-serv13460
> No space left on device
> 
> Do you have any idea what the solution to this problem could be? 

You don't have a separate filesystem for /tmp?  That's where
the space is running out.  Try the df command, it will tell you
all the filesystems you have as well as how full they are.

There are two kinds of space on unix-style filesystems:
blocks, and inodes.

You have to have enough of each in order to create new files.

The command "df" can help you see how much space you have:
Here are examples from my machine:

How many blocks?
bjb@seal ~/projects/malloc/mt90528 679 $ df
Filesystem           1k-blocks      Used Available Use% Mounted on
/dev/sdb1                98602     28913     64597  31% /
/dev/sdb6              1588133   1328612    177441  88% /usr
/dev/sdb5                98602     49477     44033  53% /var
/dev/sdb4                98602        62     93448   0% /tmp
/dev/sda1              1883573   1612967    173250  90% /home

How many inodes?
bjb@seal ~/projects/malloc/mt90528 681 $ df -i
Filesystem            Inodes   IUsed   IFree IUse% Mounted on
/dev/sdb1              25480    4207   21273   17% /
/dev/sdb6             411648   93732  317916   23% /usr
/dev/sdb5              25480    5745   19735   23% /var
/dev/sdb4              25480      27   25453    0% /tmp
/dev/sda1             487424   14741  472683    3% /home

It is often the case that one runs out before the other,
as you can see from my home directory, whose blocks are 90% used
and whose inodes are only 3% used.  You get an inode per file,
so when files are large you run out of blocks first.
When you have bazillions of small files (such as when you
run a news server) you run out of inodes first.

-- 
bjb@achilles.net
Welcome to the GNU age!   http://www.gnu.org



Reply to: