http://bugzilla.suse.com/show_bug.cgi?id=972471 Per Jessen changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |per@computer.org -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c2 --- Comment #2 from George Anchev --- Disabling IPv6 is not a solution, maybe just a workaround. I still want to be able to access Internet hosts through IPv6. Why should we disable a whole network protocol just to get a normal boot time? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c4 Pawel Wieczorkiewicz changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |pwieczorkiewicz@suse.com, | |studio@anchev.net Flags| |needinfo?(studio@anchev.net | |) --- Comment #4 from Pawel Wieczorkiewicz --- I believe that IPv6 related delay got fixed by the following commit: https://github.com/openSUSE/wicked/commit/079fa6210cbb7f5dee687aba21f50d0e19... with version 0.6.30. What wicked version are you testing on? Could you provide some logs: https://en.opensuse.org/openSUSE:Bugreport_wicked -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c6 Marius Tomaschewski changed: What |Removed |Added ---------------------------------------------------------------------------- See Also| |https://bugzilla.opensuse.o | |rg/show_bug.cgi?id=916956 --- Comment #6 from Marius Tomaschewski --- (In reply to Dave Howorth from comment #1)

I have encountered this problem on a new install of Leap 42.1 on an Acer XC-705. In my case the wicked delay went away after I disabled IPv6 in YaST.

There is still a noticeable delay that seems to be attributable to ModemManager and/or SuSE_firewall2

SuSEfirewall2 as wicked is not using ModemManager yet. Does it also happen when you disable SuSEfirewall2(_init) services? Also, try to set: /etc/sysconfig/SuSEfirewall2, FW_BOOT_FULL_INIT=yes The SuSEfirewall2_init script blocks most icmpv6 traffic including IPv6 RA used by the kernel to set default route + autoip and start dhcpv6 [managed mode] processing in wicked. FW_BOOT_FULL_INIT=yes causes to apply final rules already in the init phase, so the packets are not blocked while the setup. Except of this, the kernel forgets to send a NEWLINK when it receives RA in some cases. When there is also a prefix in the RA, wicked will workaround it, but when not, it does not get informed. See also: http://patchwork.ozlabs.org/patch/512486/ http://patchwork.ozlabs.org/patch/512646/ which still aren't applied to the kernels... -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c8 --- Comment #8 from Marius Tomaschewski --- Please disable SuSEfirewall2, enable debug, reboot and attach "wicked show-config" and "journalctl -b -o short-precise" outputs when it still happens with disabled SuSEfirewall2, see also: https://en.opensuse.org/openSUSE:Bugreport_wicked -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c9 --- Comment #9 from George Anchev --- (In reply to Marius Tomaschewski from comment #8)

Please disable SuSEfirewall2, enable debug, reboot and attach "wicked show-config" and "journalctl -b -o short-precise" outputs when it still happens with disabled SuSEfirewall2, see also:

https://en.opensuse.org/openSUSE:Bugreport_wicked

The link explains how to enable debug. How do I disable it after that and restore things back to the original (non-debugging) state? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c11 --- Comment #11 from Per Jessen --- (In reply to Marius Tomaschewski from comment #8)

Please disable SuSEfirewall2, enable debug, reboot and attach "wicked show-config" and "journalctl -b -o short-precise" outputs when it still happens with disabled SuSEfirewall2, see also:

https://en.opensuse.org/openSUSE:Bugreport_wicked

FYI, in don't use the susefirewall, but I can confirm the 30s wait as described by George and Dave. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c16 --- Comment #16 from George Anchev --- (In reply to Pawel Wieczorkiewicz from comment #15)

After upgrade to Leap the interface naming scheme changes: interfaces by default are named again as eth0, eth1, ... .

I see you still have olf config files: ifcfg-eno1 etc. Therefore a wicked client waits for $WAIT_FOR_INTERFACES (30s) timeout for all interfaces that have configs specified (even though the interface with old names are not to show up). Please verify what you interfaces names are and leave only relevant configs (or change the STARTMODE to off or manual in the old ones).

When I open YaST->Network settings I see eth0 and eth1. What are you talking about? Can you please explain how to fix the thing you explain? (sorry, I am not an expert) -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c18 --- Comment #18 from George Anchev --- (In reply to Pawel Wieczorkiewicz from comment #17)

Sure. Please inspect manually a directory /etc/sysconfig/network. Look for ifcfg-NAME files that do not correspond to the interfaces present in your system. I see in the logs you provided that your system still has: /etc/sysconfig/network/ifcfg-enp2s0 /etc/sysconfig/network/ifcfg-eno1 (remove them)

along the new ones: /etc/sysconfig/network/ifcfg-eth0 /etc/sysconfig/network/ifcfg-eth1 (which are ok and should stay)

Thanks. Done and rebooted. No change at all - still waiting 33 sec for wicked... -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c20 --- Comment #20 from George Anchev --- (In reply to Pawel Wieczorkiewicz from comment #19)

I see that eth0 has no link up event: Apr 04 12:24:43.304117 i7 wickedd-nanny[1350]: eth0: state=device-up want=network-up, wait-for=link-up

Is this interface connected to some network? If not (or there is no carrier on the cable of that network), please append: LINK_REQUIRED=no into /etc/sysconfig/network/ifcfg-eth0 and try over.

eth0 is my second interface. It is used for a second internal LAN (which is not always available and online, I turn on the network switch only when I need to share the Internet to other computers). I have not changed its configuration in the 6 months at all. So it has always been the way it is and never caused this delay. Should I still do something after this explanation? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c22 --- Comment #22 from George Anchev --- (In reply to Marius Tomaschewski from comment #21) As noted in a previous reply I have already removed eno1 and enp2s0 but that didn't change anything. As for eth1 causing the timeout: The question is why has it never caused such waiting so far and suddenly after the system update 2-3 weeks ago it is causing a problem? I don't know what is the proper configuration. I just noticed the long wait and reported. The actual network structure is: eth1 - connected to the main router (so called external) eth0 - second internal LAN which as I explained is not always online, i.e. the net switch is turned on only when the second LAN is needed Please let me know if I really need to change anything in config. As for splitting the bug - I don't know. As long as it gets fixed, it is up to you I guess. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c24 --- Comment #24 from Per Jessen --- Created attachment 671462 --> http://bugzilla.suse.com/attachment.cgi?id=671462&action=edit output from the pre-up script -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c51 Marius Tomaschewski changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(studio@anchev.net | |) --- Comment #51 from Marius Tomaschewski --- (In reply to George Anchev from comment #50)

It is in no way related to cables or anything else. As I mentioned - same settings, same hardware, same network, same cables. The only different thing is this extra 33 seconds.

OK, then please install the old *wicked* version which still worked and provide debug as described in comment 44 from both versions. A 'hwinfo --netcard | grep "Driver Modules"' will show you the module names to use in rmmod/modprobe lines instead of tg3. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c53 --- Comment #53 from George Anchev --- (In reply to Marius Tomaschewski from comment #51)

OK, then please install the old *wicked* version which still worked and provide debug as described in comment 44 from both versions.

Which one is the old? In YaST I search or wicked and see: 0.6.30-6.1-x86_64 from vendor openSUSE (Installed) 0.6.28-3.1-x86_64 from openSUSE-leap/42.1-Update with priority 99 and vendor openSUSE 0.6.27-1.1-x86_64 from Main Repository (OSS) with priority 99 and vendor openSUSE Do I have to downgrade wicked-service and libwicked-0-6 too?

A 'hwinfo --netcard | grep "Driver Modules"' will show you the module names to use in rmmod/modprobe lines instead of tg3.

This gives: Driver Modules: "e1000e" Driver Modules: "r8169" Also do I have to re-debug again for current version as I have already sent the debug for it in an earlier comment? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c56 --- Comment #56 from George Anchev --- (In reply to Marius Tomaschewski from comment #55)

(In reply to Per Jessen from comment #54)

...

(In reply to Marius Tomaschewski from comment #52)

...

Per, does the change / RPMs from comment 45 remove the dhcpv6 delays for you?

Yes, I installed yesterday, and observed a delay of about 12secs, much better. Is that about what you expected?

Yes, I'd say 12sec are OK on a cold/unused link - depends on the NIC [how long it needs to initialize, to detect carrier, forwards packets after it reported link UP] and if your router is using a "3sec silence between answers" or not (some do).

(In reply to George Anchev from comment #53)

...

(In reply to Marius Tomaschewski from comment #51) Which one is the old? In YaST I search or wicked and see:

0.6.30-6.1-x86_64 from vendor openSUSE (Installed) 0.6.28-3.1-x86_64 from openSUSE-leap/42.1-Update with priority 99 and vendor openSUSE 0.6.27-1.1-x86_64 from Main Repository (OSS) with priority 99 and vendor openSUSE

I've no idea which version were working for you before. See /var/log/zypp/history, it should show what were installed when.

...

Do I have to downgrade wicked-service and libwicked-0-6 too?

Yes please. It should be not even possible to omit them.

...

Also do I have to re-debug again for current version as I have already sent the debug for it in an earlier comment?

I'd say not the current (old 0.6.30 in the meantime), but master build:

http://download.opensuse.org/repositories/network:/wicked:/master/ openSUSE_Leap_42.1/

which contains the missed kernel event workaround fix from comment 46.

Ok, I have tried all 3 versions: 0.6.30 0.6.31 (from master) 0.6.28 (which I found in the zypper history log) - this one I even tried with reverting to the previous version of systemd With all these the delay is still 33 seconds. Bummer. I also tried to set eth0 to "On cable connection" and to "Never". Only when set to boot "Never" the delay is gone. I followed the procedure from comment 44. I didn't reboot between creating the logs. The exact thing I did was: 1. Go to YaST and install the package versions 2. Execute commands: rcwickedd restart tcpdump -envfi any icmp6 2>&1 | systemd-cat -t "tcpdump" & ip monitor link address prefix neigh route | systemd-cat -t "ip-monitor" & rmmod r8169 rmmod e1000e cursor=`journalctl -n 10 -o export | grep ^__CURSOR= | sed -e 's|^__CURSOR=||' | head -1` modprobe r8169 modprobe e1000e sleep 40 ip a s | systemd-cat -t "ip-addrs" ip -6 r s | systemd-cat -t "ip-route6" sysctl -a | grep ^net.ipv6.conf | systemd-cat -t "ipv6-sysctls" kill $(jobs -lp) journalctl -c "$cursor" -o short-precise > journal-XXX.txt where XXX is the version of wicked as you will see in the 3 log files: https://goo.gl/mnSuJx Another thing I would like to ask for further clarification: (In reply to Marius Tomaschewski from comment #49)

(In reply to George Anchev from comment #48)

...

Also - if I lower manually WAIT_FOR_INTERACES to (say) 3 seconds - what could be the negative impact?

ifup will return after 3 sec, setup will happen in background and when you have services which bind to specific IPs, the'll fail.

How do I check which these services are and if they bind to specific IPs? The internal zone interface (eth0) is used for sharing the internet connection to that internal network + file transfers through sftp and smb. All I need is to make sure that when that network is online (the net switch is turned on) it will be possible to use these network functions + of course get rid of the 33 sec delay. I hope you can help. Thanks. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 Sagi Ben Akiva changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |sagiben@gmail.com -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c64 --- Comment #64 from Bernhard Wiedemann --- This is an autogenerated message for OBS integration: This bug (972471) was mentioned in https://build.opensuse.org/request/show/440828 13.2 / wicked https://build.opensuse.org/request/show/440829 42.1 / wicked https://build.opensuse.org/request/show/440830 42.2 / wicked -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 Swamp Workflow Management changed: What |Removed |Added ---------------------------------------------------------------------------- Whiteboard|maint:planned:update |maint:planned:update |ibs:running:3624:moderate |ibs:running:3624:moderate |ibs:running:3623:moderate |ibs:running:3623:moderate | |obs:running:5911:moderate -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c66 Marius Tomaschewski changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |bugrprt21882@online.de --- Comment #66 from Marius Tomaschewski --- *** Bug 993516 has been marked as a duplicate of this bug. *** -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c67 --- Comment #67 from Swamp Workflow Management --- openSUSE-RU-2016:2978-1: An update that has 19 recommended fixes can now be installed. Category: recommended (moderate) Bug References: 865573,902897,913861,916613,916948,972471,973355,974231,975466,981887,982231,982952,984088,985894,988794,988954,989741,997027,998413 CVE References: Sources used: openSUSE Leap 42.1 (src): wicked-0.6.39-12.1 openSUSE 13.2 (src): wicked-0.6.39-36.1 -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 http://bugzilla.suse.com/show_bug.cgi?id=972471#c69 --- Comment #69 from Swamp Workflow Management --- SUSE-RU-2016:3055-1: An update that has 6 recommended fixes can now be installed. Category: recommended (moderate) Bug References: 972471,975466,988794,988954,997027,998413 CVE References: Sources used: SUSE Linux Enterprise Server for Raspberry Pi 12-SP2 (src): wicked-0.6.39-34.1 SUSE Linux Enterprise Server 12-SP2 (src): wicked-0.6.39-34.1 SUSE Linux Enterprise Desktop 12-SP2 (src): wicked-0.6.39-34.1 -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 Swamp Workflow Management changed: What |Removed |Added ---------------------------------------------------------------------------- Whiteboard|maint:planned:update |maint:planned:update |ibs:running:3624:moderate |ibs:running:3624:moderate |ibs:running:3623:moderate |obs:running:6033:moderate |obs:running:6033:moderate | -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 Swamp Workflow Management changed: What |Removed |Added ---------------------------------------------------------------------------- Whiteboard|maint:planned:update |maint:planned:update |ibs:running:3624:moderate |ibs:running:3624:moderate |obs:running:6033:moderate | -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 Swamp Workflow Management changed: What |Removed |Added ---------------------------------------------------------------------------- Whiteboard|maint:planned:update |maint:planned:update |ibs:running:3624:moderate | -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.suse.com/show_bug.cgi?id=972471 Swamp Workflow Management changed: What |Removed |Added ---------------------------------------------------------------------------- Whiteboard|maint:planned:update | -- You are receiving this mail because: You are on the CC list for the bug.