About this list Date view Thread view Subject view Author view Attachment view

From: Herbert Poetzl (herbert_at_13thfloor.at)
Date: Thu 04 Sep 2003 - 04:41:58 BST


On Thu, Sep 04, 2003 at 11:54:38AM +0100, Luís Miguel Silva wrote:
> I had exactly the same problem and was to lazy to search for an answer (so
> i just kept using kernel 2.4.20...)
> this happened till a friend of mine bothered to go to google and find some
> info on this mailing list ;o)
>
> So...(by a big coincidence)...right now, its 3:40am in Portugal and im
> upgrading all our production servers to kernel 2.4.21+ctx17 (with patches
> from herberts site).
>
> Has Solucorps site changed or..was the following text on the site right
> from the start:
> patch-2.4.20ctx-17.gz - The patch against Linux 2.4.20
> patch-2.4.21ctx-17.gz - A ->test<- patch against Linux 2.4.21
>
> I downloaded this patch from the site about 6 times and it didnt caught my
> eye that it was a TEST patch for linux 2.4.21! (silly and lazy me...).
>
> By the way, i just tried to use the patches against 2.4.22 from herberts
> site...and, allthough they compiled just fine, vservers couldnt boot!
>
> Im too sleepy right now to tell you the exact error, but, it was something
> "about not being able to set up the ulimits"...

well, the ulimit issue comes with the changed behaviour of
linux 2.4.22 vs. 2.4.21 and has nothing to do with vserver,
except for the fact that vserver scripts use ulimit and the
scripts didn't change ...

if you take a look at my patchset
http://www.13thfloor.at/VServer/patches-2.4.22-c17/
you'll find a ulimit-unfix patch, which reverts this
kernel behaviour ...

make sure that, if you use ctx17a (which would be advised)
that you also have the 0.23 vserver tools, which in turn
will give another warning
"Kernel do not support chrootsafe(), using chroot()", which
can be safely ignored, because it only states that the ctx17a
patch doesn't support chrootsafe yet ...

> Another big curiosity (allthough im not sure if the vservers patch as
> ANYTHING to do with it (probably not)), since i upgraded the kernel to
> 2.4.21 ctx17 (just some minutes ago), my primary RAID disk (which has the
> system and all the vservers) suddenly downgraded to about 2 weeks ago by
> assuming old files (heh?...long live real'time'mirrors by raid!)
>
> Ever happened that to any of you? :oP
>
> Well..oh my..i guess ill just have to face the ungry users in the morning!
> ;o)
>
> PS: excuse my *bad* english. IM LAZY AND SLEEPY! :oP

you are excused ;)

HTH,
Herbert

> Regards,
> +-----------------------------------------
> | Luís Miguel Silva
> | Network Administrator@ ISPGaya.pt
> | Rua António Rodrigues da Rocha, 291/341
> | Sto. Ovídio • 4400-025 V. N. de Gaia
> | Portugal
> | T: +351 22 3745730/3/5 F: +351 22 3745738
> | G: +351 93 6371253 E: lms_at_ispgaya.pt
> | H: http://lms.ispgaya.pt/
> +-----------------------------------------
> > Hello,
> >
> > I was compilling vanilla kernel patched with ctx-17a patch, I was not
> > able to compile code udp.c if ipv6 support was not selected.
> >
> > I was searching and I found bug in file linux/net/ipv4/udp.c.
> >
> > It was not clear to me what patch author intended...
> > but there is for sure a bug.
> >
> > look for "#if defined(CONFIG_IPV6)" in the code:
> >
> > Michal Heppler
> >
> > --
> >
> > /* UDP is nearly always wildcards out the wazoo, it makes no sense to
> > * try
> > * harder than this. -DaveM
> > */
> > struct sock *udp_v4_lookup_longway(u32 saddr, u16 sport, u32 daddr, u16
> > dport, int dif)
> > {
> > struct sock *sk, *result = NULL;
> > unsigned short hnum = ntohs(dport);
> > int badness = -1;
> >
> > for(sk = udp_hash[hnum & (UDP_HTABLE_SIZE - 1)]; sk != NULL; sk
> > = sk->next) {
> > if(sk->num == hnum && !ipv6_only_sock(sk)) {
> > int score;
> > #if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
> > score = sk->family == PF_INET ? 1 : 0;
> > #else
> > }else if (sk->ip_info != NULL){
> > if (udp_in_list (sk->ip_info,daddr)){
> > score++;
> > }else{
> > continue;
> > }
> > score = 1;
> > #endif
> > if(sk->rcv_saddr) {
> > if(sk->rcv_saddr != daddr)
> > continue;
> > score+=2;
> >
> > ,,,
> >
> >
> > }
> > }
> > }
> > return result;
> > }
>
>


About this list Date view Thread view Subject view Author view Attachment view
[Next/Previous Months] [Main vserver Project Homepage] [Howto Subscribe/Unsubscribe] [Paul Sladen's vserver stuff]
Generated on Thu 04 Sep 2003 - 05:04:19 BST by hypermail 2.1.3