Subject: out of vnodes
To: None <port-pc532@netbsd.org>
From: Jon Buller <jonb@metronet.com>
List: port-pc532
Date: 01/26/1999 08:25:42
I cant link libc right now, as I get messages that look like:

Jan 26 07:15:14 bullbox /netbsd: vnode: table is full
Jan 26 07:15:14 bullbox syslogd: /var/run/utmp: Too many open files in system
Jan 26 07:15:15 bullbox last message repeated 11 times
Jan 26 07:15:15 bullbox /netbsd: 3>vnode: table is full
Jan 26 07:15:15 bullbox /netbsd: vnode: table is full
Jan 26 07:15:15 bullbox last message repeated 18 times
Jan 26 07:15:18 bullbox last message repeated 318 times
Jan 26 07:16:03 bullbox last message repeated 2 times
Jan 26 07:34:12 bullbox /netbsd: vnode: table is full
Jan 26 07:34:12 bullbox syslogd: /var/run/utmp: Too many open files in system
Jan 26 07:34:13 bullbox last message repeated 11 times
Jan 26 07:34:13 bullbox /netbsd: 3>vnode: table is full
Jan 26 07:34:13 bullbox /netbsd: vnode: table is full
Jan 26 07:34:16 bullbox last message repeated 314 times
Jan 26 07:35:01 bullbox last message repeated 2 times
Jan 26 07:34:13 bullbox last message repeated 18 times
Jan 26 07:43:17 bullbox last message repeated 192 times

(I tried turning off apache, the NFS daemons, postgresql, etc.  I didn't
think it would be enough, but I thought I'd try...)

Is my new linker broken?  Can I correct this with "options NVNODES=xxx"
for some xxx in my kernel config?  Any suggestions for what xxx
should be?  If that's not it, should I just bump maxusers from 8
to 16?  Something else?

Thanks in Advance,
Jon