Roman Drahtmueller sagte:
We fix remaining bugs. There are some that we don't want people to experience, such as IDE hangs
IDE hangs?!?!?
maybe THAT is what kills my SuSE 8.0 all the time?
Got a promise controller and a samsung 80Gig disk in my firewall/server. whenever my main box is in linux, suddenly everything stops. And I _DO_ mean everything; both machines freeze in the very same moment. As long as my main box is on windows everything's fine.
By now I think I traced it down to that ist started to happen shortly after I applied the glibc patch available in YOU...
I'll reverse that one tonite and have another go at it.
The glibc update has nothing to do with it, if the whole machine hangs, then it must be a kernel problem.
strange facts though: 1. the box has been running the same kernel version on that hardware for several months by now, without any problems. 2. it started shortly after I applied the glibc patch. 3. it started to happen more frequent when I applied the same patch on my main workstation as well. 4. there's nothing in all the logfiles that shows any hardware related error message, or kernel message. 5. as long as the box is 'alone' nfs-wise I can throw load at it as I want... stable as rock. As soon as I login on my other box where the account is a NIS account with a home directory served by NFS from the box in question, it gets risky. If my main box is on windoze, the server is good for another week of uptime, no matter which load. 6. doesn't matter wether I run the kernel nfs server or the userspace nfs server.
Please try Hubert's kernel RPMs from ftp://ftp.leo.org/pub/people/mantel/next/RPM/ and see if it vanishes. Do not forget mk_initrd and lilo.
sure will do; anything to get rid of those hangs bye, [MH]