Re: Corrupted IPv6 routing table

2015-01-08 Thread Przemyslaw Frasunek
>> Has anyone seen this before? > Fixed in 9 by r257389 (So you should try either stable or 10.x). Thanks a lot! I'll schedule an upgrade. ___ freebsd-net@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, se

Corrupted IPv6 routing table

2015-01-07 Thread Przemyslaw Frasunek
Dear all, We are running FreeBSD 9.2-RELEASE-p3 on few PPPoE access servers, each servicing about 1000 customers. Each server exchanges customers' /32 (for IPv4) and /64 (for IPv6) routes using OSPF and BIRD. Few times in a month, we are experiencing routing table corruption, which causes spuriou

Re: mpd5/Netgraph issues after upgrading to 7.4

2014-04-20 Thread Przemyslaw Frasunek
FYI -- after upgrade to 9-STABLE no further crashes occurred, even with IPv6 enabled. >>> What sort of uptime have you seen with ipv6 enabled ? >> Now it's 19 days and still no crash occurred. > Has all been stable still with ipv6 enabled ? Unfortunately, it crashed after ~30 d

Re: mpd5/Netgraph issues after upgrading to 7.4

2014-03-14 Thread Przemyslaw Frasunek
>> FYI -- after upgrade to 9-STABLE no further crashes occurred, even with IPv6 >> enabled. > What sort of uptime have you seen with ipv6 enabled ? Now it's 19 days and still no crash occurred. ___ freebsd-net@freebsd.org mailing list http://lists.freeb

Re: mpd5/Netgraph issues after upgrading to 7.4

2014-03-09 Thread Przemyslaw Frasunek
>> I've seen that Mike reported similar issues in October >> (http://lists.freebsd.org/pipermail/freebsd-stable/2013-October/075552.html). >> Did you managed to resolve it? > I worked around the crash by removing ipv6 from the kernel. The box has > been functioning without a crash since then.

Re: mpd5/Netgraph issues after upgrading to 7.4

2012-07-09 Thread Przemyslaw Frasunek
> It seems, Przemyslaw Frasunek uses proxyarp? > I have no such problems but I do not use proxyarp. > Could you get rid of it, Przemyslaw? No, I don't use proxy ARP. I have about 300 PPPoE ng interfaces and 10 VLANs with plain IP traffic. ARP table has only < 50 entries, all of

Re: mpd5/Netgraph issues after upgrading to 7.4

2012-07-07 Thread Przemyslaw Frasunek
> Did you set net.isr.direct=0 (and/or direct_force)? > If so, don't do that. Get back to default 1 for these two sysctls. Both sysctls are set to default values. This is my /etc/sysctl.conf: net.inet6.ip6.redirect=0 net.inet.icmp.drop_redirect=1 net.inet6.icmp6.rediraccept=0 hw.acpi.power_butto

Re: mpd5/Netgraph issues after upgrading to 7.4

2012-07-07 Thread Przemyslaw Frasunek
> After reenabling IPv6, the crash occurred within 6 hours. This time, crashdump > was properly saved (thanks to patch suggested by Eugene). My PPPoE BRAS was stable for 17 days. This morning, it crashed in another way: current process = 2762 (mpd5) trap number = 9 panic: gene

Re: mpd5/Netgraph issues after upgrading to 7.4

2012-06-19 Thread Przemyslaw Frasunek
> It's way better now. I had no crash since IPv6 is disabled. After reenabling IPv6, the crash occurred within 6 hours. This time, crashdump was properly saved (thanks to patch suggested by Eugene). As already stated by bz, panic is definitely related to races in IPv6 code: (kgdb) bt #0 doadump

Re: mpd5/Netgraph issues after upgrading to 7.4

2012-06-18 Thread Przemyslaw Frasunek
>> Thanks a lot guys. For now, I disabled IPv6 on this BRAS. Let's see if it's >> going to help. > Hi, > Any changes in stability ? Hi, It's way better now. I had no crash since IPv6 is disabled. ___ freebsd-net@freebsd.org mailing list http://li

Re: mpd5/Netgraph issues after upgrading to 7.4

2012-06-15 Thread Przemyslaw Frasunek
> I suspect this isn't related to netgraph, but to IPv6 since prelist_remove() > is found in netinet6/nd6_rtr.c. > > Several times I looked into ND code and found lots of race prone code there. > May be some was recently fixed by bz@, but definitely not merged to stable/8. Thanks a lot guys. For

Re: mpd5/Netgraph issues after upgrading to 7.4

2012-06-15 Thread Przemyslaw Frasunek
> One more: does your box has PS/2 keyboard or USB? It matters too. > For systems having USB keyboard there is another patch needed to obtain > crashdumps (by Andriy Gapon): Thanks a lot. I have KVM connected using USB. I'll apply this patch. ___ freebs

Re: mpd5/Netgraph issues after upgrading to 7.4

2012-06-15 Thread Przemyslaw Frasunek
Dear All, unfortunately, one of my mpd5 PPPoE access servers started panicing every few hours. I'm running recent 8.3-STABLE (as of 23th May) with WITNESS, INVARIANTS and DEBUG_MEMGUARD compiled. Unfortunately, I'm unable to catch crashdump. For some reason, it is not saved on dumpdev. The only

dummynet damages ICMPv6 packets

2011-11-07 Thread Przemyslaw Frasunek
Hello, we are experiencing interesing behaviour of dummynet enabled on IPv6 interfaces. When the following rules are added: add pipe 24 ip from any to any in recv vlan1 add pipe 25 ip from any to any out xmit vlan1 all ICMPv6 packets passing on vlan1 are being damaged: 10:55:53.180801 IP6 fe80:

Re: kern/122290: [netgraph] [panic] Netgraph related "kmem_map too small" panics

2011-05-10 Thread Przemyslaw Frasunek
The following reply was made to PR kern/122290; it has been noted by GNATS. From: Przemyslaw Frasunek To: bug-follo...@freebsd.org Cc: Subject: Re: kern/122290: [netgraph] [panic] Netgraph related "kmem_map too small" panics Date: Tue, 10 May 2011 22:30:31 +0200 It seems to be f

Re: mpd5/Netgraph issues after upgrading to 7.4

2011-04-11 Thread Przemyslaw Frasunek
> T> Ah, I found where EPERM comes from. Will fix it soon. > Sorry, I was wrong :( I managed to track down EPERMs back to ng_make_node() call from ng_mkpeer() in netgraph/ng_base.c:1460. I'll add some additional printfs inside ng_make_node(). ___ freebs

Re: mpd5/Netgraph issues after upgrading to 7.4

2011-04-11 Thread Przemyslaw Frasunek
> Do you use i386 or amd64? I use amd64, it has much more KVA space > (kernel virtual area). I use i386. I suppose I'll need to raise KVA_PAGES to 512 and then slightly raise vm.kmem_size. What are the safe values for i386? I wonder why I got crash without dump and not the "kmem_map too small" pa

Re: mpd5/Netgraph issues after upgrading to 7.4

2011-04-11 Thread Przemyslaw Frasunek
> Increase sysctl kern.ipc.maxsockbuf. > I was forced to rise it upto 80MB (sic!) as 8MB was not enough to me. Yay, things are getting worse. Increasing maxsockbuf caused crash after 2-3 hours. There was no crashdump and the last thing in log was: Apr 11 12:32:40 lsm-gw kernel: ad4: FAILURE - out

Re: mpd5/Netgraph issues after upgrading to 7.4

2011-04-10 Thread Przemyslaw Frasunek
> IMO, any kind of memory allocation code (malloc, uma, netgraph item > allocator) never return EPERM, they return ENOMEM or ENOBUFS. > > So, there is a bug somewhere else. I think so, but for me it still looks like resource shortage. As I wrote before, when EPERM starts appearing, I'm unable to

Re: mpd5/Netgraph issues after upgrading to 7.4

2011-04-10 Thread Przemyslaw Frasunek
> Use command "vmstat -z|egrep 'ITEM|NetGraph'" and check FAILURES column. > If you see non-zero values there, you need to increase netgraph memory limits > net.graph.maxdata and net.graph.maxalloc using /boot/loader.conf. Unfortunately, increasing net.graph.maxdata & net.graph.maxalloc didn't sol

Re: mpd5/Netgraph issues after upgrading to 7.4

2011-04-10 Thread Przemyslaw Frasunek
> Use command "vmstat -z|egrep 'ITEM|NetGraph'" and check FAILURES column. > If you see non-zero values there, you need to increase netgraph memory limits > net.graph.maxdata and net.graph.maxalloc using /boot/loader.conf. Thanks, indeed it helped a lot. I also noticed that other zones have non-ze

Re: mpd5/Netgraph issues after upgrading to 7.4

2011-04-10 Thread Przemyslaw Frasunek
W dniu 08.04.2011 22:13, Przemyslaw Frasunek pisze: > I'm still looking for a help in investigating this issue. The problem appears > on > two 7.4 boxes, while 7.3 are working OK. Ktrace shows, that indeed some of > sendto() calls on netgraph control socket are f

Re: mpd5/Netgraph issues after upgrading to 7.4

2011-04-08 Thread Przemyslaw Frasunek
[...] > Mar 31 13:48:06 lsm-gw mpd: [B-150] Bundle: Interface ng149 created > Mar 31 13:48:06 lsm-gw mpd: [B-150] can't create ppp node at ".:"->"b150": > Operation not permitted [...] I'm still looking for a help in investigating this issue. The problem appears on two 7.4 boxes, while 7.3 are wor

Re: panic: bufwrite: buffer is not busy???

2011-04-08 Thread Przemyslaw Frasunek
> I believe this netgraph/mpd stability problem it solved by glebius' patches. > I have been testing them in lab and in production for many weeks and they just > eliminated my panics altogether. Those patched have been commited to HEAD and > RELENG_8. Any chance of a backporting them to RELENG_7?

mpd5/Netgraph issues after upgrading to 7.4

2011-03-31 Thread Przemyslaw Frasunek
Hello, I have upgraded one of my mpd5 based PPPoE access servers from 7.3-RELEASE to 7.4-RELEASE. Just after upgrade, I started getting following errors: Mar 31 13:48:06 lsm-gw mpd: [B-150] Bundle: Interface ng149 created Mar 31 13:48:06 lsm-gw mpd: [B-150] can't create ppp node at ".:"->"b150":

Re: Netgraph/mpd5 stability issues

2011-02-16 Thread Przemyslaw Frasunek
> On 14.01.2011 18:46, Mike Tancsa wrote: > I also have very loaded mpd/PPPoE servers that panic all the time: > > http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/153255 > http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/153671 I've just got yet another panic on mpd5 box after about 30 days of upt

Re: Netgraph/mpd5 stability issues

2011-01-31 Thread Przemyslaw Frasunek
> And in this one, can you please show *hook->hk_peer ? (kgdb) print *hook->hk_peer $2 = { hk_name = "\b\000\000\000 \000\000\000\004\000\000\000\001\000\000\000ŐRí\003\003ö\0248cmd4\000\000\000", hk_private = 0x0, hk_flags = 0, hk_refs = 0, hk_type = 0, hk_peer = 0x0, hk_node = 0x0, hk_hooks

Re: Netgraph/mpd5 stability issues

2011-01-31 Thread Przemyslaw Frasunek
> (kgdb) print *ng_name_hash[116].lh_first It looks like this one is corrupted: (kgdb) print *ng_name_hash[116].lh_first.nd_nodes.le_next.nd_nodes.le_next.nd_nodes.le_next.nd_nodes.le_next $19 = {nd_name = "ng258", '\0' , nd_type = 0xc61871a0, nd_flags = 0, nd_refs = 1, nd_numhooks = 0, nd_priv

Re: Netgraph/mpd5 stability issues

2011-01-31 Thread Przemyslaw Frasunek
> In this dump, can we seek for where did 0x74 came from? Can you look at > ng_name_hash[hash]? (kgdb) print hash No symbol "hash" in current context. (kgdb) info all eax0xff9a -102 ecx0xe7ce6895 -405903211 edx0xff9a -102 ebx

Netgraph/mpd5 stability issues

2011-01-14 Thread Przemyslaw Frasunek
Hello, I'm using mpd 5.5 on three PPPoE routers, each servicing about 300 PPPoE concurrent sessions. Routers are based on Intel SR1630GP hardware platforms and runs FreeBSD 7.3-RELEASE. I'm experiencing stability issues related to Netgraph. None of above routers can survive more than 20-30 days o