Stephen
On 14 Oct 1999, Jarno Paananen wrote:
> Hi,
>
> the problem is that the Unix98 PTY numbers don't seem to get
> reused. With a maximum of 256 configured in the kernel it
> takes 1-2 weeks to run out of PTYs.
>
> At the moment the situation is like this (the machine had to be
> booted yesterday because of this and now it seems we have to do
> it again before end of the week):
>
> crw------- 1 chery tty 136, 0 Oct 13 21:34 0
> crw--w---- 1 pekangas tty 136, 1 Oct 13 23:59 1
> crw------- 1 jpaana tty 136, 104 Oct 14 00:00 104
> crwx------ 1 jpaana tty 136, 108 Oct 14 00:00 108
> crw------- 1 jpaana tty 136, 3 Oct 13 23:59 3
> crw------- 1 jpaana tty 136, 4 Oct 14 00:00 4
> crw------- 1 chery tty 136, 43 Oct 13 23:59 43
> crw------- 1 chery tty 136, 44 Oct 13 21:34 44
> crw------- 1 jpaana tty 136, 5 Oct 13 23:59 5
> crw------- 1 jpaana tty 136, 6 Oct 14 00:00 6
> crw------- 1 jpaana tty 136, 75 Oct 14 00:00 75
> crw------- 1 jpaana tty 136, 95 Oct 13 23:53 95
>
> And this is all in /dev/pts. Lsof doesn't reveal any extra open
> PTYs. After number 255 is opened all programs complain about not
> getting a PTY.
>
> The kernel version is 2.2.13pre14 compiled with gcc 2.7.2.3. The
> same symptom appeared in 2.2.12, but possibly not in 2.2.10 (at
> least that kernel got an uptime of 62 days).
>
> Could this be a userland problem with something taking up PTYs,
> a glibc bug (using 2.1.2 from Debian Potato), or a genuine
> kernel problem?
>
> Any help is very welcome...
>
> // Jarno
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.rutgers.edu
> Please read the FAQ at http://www.tux.org/lkml/
>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/