Am Mittwoch, 8. Dezember 2004 13:07 schrieb Ralf Gesel|ensetter: > I got the feeling that our tjener is to blame for peridical hangs I > observed at thin clients, and work stations. LTSPserver reports > someting like: NFS not reachable (time-out) / NFS connected ok.
Hi there, I am very close to the solution, now: - ldap/nfs overload is due to crashing nscd on ltsp. - these crashes are now compensated by starting nscd via cronjob every 5 min. - nscd depends only on libc6 - as opposed to the original skolelinux version 2.2.3, installed on my ltsp I found libc6 (2.3.2.ds1-18) > apt-cache policy libc6 > libc6: > Installed: 2.3.2.ds1-18 > Candidate: 2.3.2.ds1-18 > Version Table: > *** 2.3.2.ds1-18 0 > 100 /var/lib/dpkg/status > 2.2.5-14.3 0 > 500 http://www.linex.org woody/linex Packages > 2.2.5-11.5 0 > 500 http://security.debian.org stable/updates/main Packages > 500 ftp://ftp.skolelinux.no woody/main Packages > 500 ftp://ftp.debian.de woody/main Packages I don't know exactly where I got it from (downgrading would remove nearly everything!), but "gg: 2.3.2.ds1-18 libc6" gives me segfault reports as first hit! > Am Wochenende ber ein libc6-Upgrade in Sarge gestolpert: von libc6 > 2.3.2.ds1-16 auf libc6 2.3.2.ds1-18 und ssh antwortet nicht mehr. Sp r > dann, als ich den Rechner wieder in greifbarer N hatte, bemerke ich, > dass einige Programme mit segmentation fault beendet wurden. > Nach einem Downgrade auf die alte libc6 (die noch > in /var/cache/apt/archives lag) ist zum Gl ck alles wieder beim alten. Furthermore it seems to be from unstable (!) Regards Ralf
