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

Re: [K12OSN] Random Program crashes



On Tue, 30 Dec 2003, norbert wrote:
> Let me save you some time ....... I've tried 2.4.22, 2.4.23, 2.4.25 on
> three identical servers ... not one behaves the same !!! 2.4.23 seems to
> deal with GCompris - so far 10 hours up but Netscape &  OO still crash,
> on 2.4.25 users cannot login (thin clients) the workstations start & get
> a login screen but when a user logs in the session "disappears" and the
> workstation returns to the login screen ??????
> Again the log files have no error messages, just the normal dhcp request
> from ws00x, session opened, session closed ..... is there a 2.5 ??? :-D
> The strangest is that although I can ping from one server to the others
> & the workstations and get internet access on the workstations two of
> the servers cannot get access?
> My setup is a modem-cable into my first server & out to the  switch
> (server 1 has its own subnet 192.168.0.0 & clients)
> Server 2 has an "in" from switch1 & an out to another switch2 with its
> subnet 192.168.100.0, the gateway  is properly defined as Server 1
> Server 3 has an "in" from switch1 & an out to another switch3 with its
> subnet 192.168.200.0, the gateway  is properly defined as Server 1
>
> So before I dive into an abyss of unknowns ........ I wanted to see if
> anyone else has taken the dive :-)
>
Norbert,
	you may remember my posts from about a month ago about mysterious
crashes of ibm dual piii boxes after upgrade to 4.beta17. it was not the
kernel - i went back to the pre-upgrade kernel and the boxes still
crashed. the weird thing is i have an identical box running without any
problems with v4. the only difference is that it is very lightly loaded
and that the gigabit pro adapter is on eth1 and not on eth0 as on the
crashing boxes. those boxes have been happily running v2.x and 3.x for a
long while. there *must* be something more in v4 that we are overlooking.
i am suspecting the libraries, the problem is detection - the boxes would
hang without leaving any error messages. julius




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