Re: RH kernel 2.4.18-10 and patches from http://people.redhat.com/~mingo/TUX-patches/
J Sloan <[email protected]>
| Newsgroups | gmane.network.tux |
|---|---|
| Message-ID | <[email protected]> |
The stock kernel may not be the best choice - For good tux results, try the latest -aa kernel. This includes various performance enhancements, including O(1), low latency core and high-end VM fixes - and always includes the TuX server. Not to stir up trouble, but the tux that ships with SuSE 8.1 seems to be pretty solid :-) Joe Alex Kramarov wrote: >----- Original Message ----- >From: "Pierre POMES" <[email protected]> >To: <[email protected]> >Sent: Thursday, October 17, 2002 12:57 PM >Subject: RH kernel 2.4.18-10 and patches from >http://people.redhat.com/~mingo/TUX-patches/ > > > > >>Hi, >> >>I am currently using tux in production on a redhat 7.3 with redhat >> >> >updates, and tux > > >>2.2.5 (rpm version). I have some hangs, so I want to upgrade tux to the >> >> >lastest > > >>version, tux 2.2.7. >> >>The kernel patch tux2-full-2.4.18-final-A3 should apply on a valilla >> >> >2.4.18 kernel, > > >>but since I have a redhat kernel (2.4.18-10 source kernel, rpm format), I >> >> >thought the > > >>file tux2-full-2.4.18-final-A3 could not be applied at all. I wanted to >> >> >try, just to > > >>be sure, and a few patches applied : >> >>patching file linux/mm/filemap.c >>Hunk #1 succeeded at 137 (offset 3 lines). >> >>patching file linux/net/ipv4/route.c >>Hunk #2 succeeded at 1417 (offset 14 lines). >>Hunk #4 succeeded at 1817 (offset 14 lines). >>Hunk #6 succeeded at 1872 (offset 14 lines). >> >>All other patches are already applied. Do you think I must recompile my >> >> >redhat kernel > > >>with this four patches to get a stable setup ? Or I should rather use a >> >> >standard > > >>redhat kernel 2.4.18-10 ? >> >> > >if you really want a stable setup, i would recommend using a plain kernel >from kernel.org . anyway, you are going where noone has gone before. when i >tried to compile rdhat kernel with latest tux, i got a plain kernel, patched >it with tux, and copied the files from /net/tux to the redhat kernel, and >the 2 inclide files /include/net/tux.h and /include/net/tux_u.h .after that >i only taken the compiled tux.o module and copied it over the original tux >module that redhat supplied. The setup was stable, worked great. the problem >that i can't compile the recent redhat kernel on smp at all with the default >configuration they provide, so i can't use this techique. If you want to run >your own kernel, i would again advise you to use the plain kernel from >kernel.org. > >yea, and don't enable logging of referrers - right now it is proving to be a >reason of crashes. for other list users, my machine is running for 35 hours >with logging on (no referrer logging), stable. will leave it as is for 2 >more days, and then turn on referrer logging to see it crash, to be sure >that this is the problem. > > > >>Second question : when I subscribed to the list, I received the following >> >> >message : > > >>"To alpha-test the latest development versions of TUX, please visit >>http://people.redhat.com/~mingo/TUX-patches/ for all the latest >> >> >bleeding-edge patches > > >>and tarballs" >> >> > >this banner is a "little" out of date ... > > > >>does it mean tux 2.2.5/2.2.7 should not used in a production environnement >> >> >? I > > >>currently have some hangs with 2.2.5 version, with the same behaviour than >> >> >described > > >>in the thread 'random system crashes' (initiated by Marco Rothley, Octy 15 >> >> >2002). > >and please read my recent posting about tux version numbering ... > > > >>Thanks for your help, >> >>-- >>Pierre POMES mailto:[email protected] >>Interface Technologies >> >> >> >>_______________________________________________ >>tux-list mailing list >>[email protected] >>https://listman.redhat.com/mailman/listinfo/tux-list >> >>. >> >> > > > >_______________________________________________ >tux-list mailing list >[email protected] >https://listman.redhat.com/mailman/listinfo/tux-list > > >