[opensuse] wickedd[3079] messages
Hello, I have been trying to keep an eye on my wicked network configuration. In my log files I have numerous instances of: wickedd[3079]: wlp11s0: state changed COMPLETED -> GROUP_HANDSHAKE wickedd[3079]: wlp11s0: state changed GROUP_HANDSHAKE -> COMPLETEDwickedd[3079]: __ni_netdev_event(wlp11s0, idx=3, link-up) wickedd[3079]: sending device event "linkUp" for /org/opensuse/Network/Interface/3 Any input on what these might be, and what state is changing any why? Thanks, Don -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 04/25/2015 09:59 AM, don fisher wrote:
Hello,
I have been trying to keep an eye on my wicked network configuration. In my log files I have numerous instances of:
wickedd[3079]: wlp11s0: state changed COMPLETED -> GROUP_HANDSHAKE wickedd[3079]: wlp11s0: state changed GROUP_HANDSHAKE -> COMPLETEDwickedd[3079]: __ni_netdev_event(wlp11s0, idx=3, link-up) wickedd[3079]: sending device event "linkUp" for /org/opensuse/Network/Interface/3
Any input on what these might be, and what state is changing any why?
Thanks, Don
I am one of the users with both wifi and cable in their laptop, so 2 devices get configured. I did and ifdown on the cable link and received about 40 wickedd[3079] messages shutting down the cable link. So the [3079] message appears to be a "for information" type message. The above wlp11s0 messages are repeated every hour at XX:28:29. It continues even after the ifdown on the cable connection. I do not know what state changed, but evidently it is not significant. Don -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 04/25/2015 02:53 PM, don fisher wrote:
So the [3079] message appears to be a "for information" type message.
Indeed. The "[3079]" is the process number doing the reporting. But the 'for info is actually a level of the hierarchy. You can over-ride that. Its back to RTFM, as always. See http://rpm.pbone.net/index.php3/stat/45/idpl/27311027/numer/8/nazwa/wickedd <quote> --log-level level Set log level to one of<error|warning|notice|info|debug>. --log-target target Set log target to one of <stderr|syslog>,optionally followed by a colon and target specific details. stderr[:options]with the following options:pidinclude program pid in each message syslog[:facility[:options]]with following facilities:user, daemon, local0 .. Local7and options:perrorlog the message to stderr as well </quote> $ man 2 syslog or $ man 3 syslog might be illuminating. Failing that, there's wikipedia http://en.wikipedia.org/wiki/Syslog#Severity_levels -- A: Yes. > Q: Are you sure? >> A: Because it reverses the logical flow of conversation. >>> Q: Why is top posting frowned upon? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
man 3 syslog Thanks. I did not recognize the PID. Do you know how to start klogd? I
On 04/25/2015 12:24 PM, Anton Aylward wrote: tried the usual: sudo systemctl enable klogd.service sudo systemctl start klogd.service but received an error message: Failed to start klogd.service: Operation refused, unit klogd.service may be requested by dependency only. I know what a dependency is, just not what the message implies. I may have some hardware problems and was trying to get more advanced reporting on segfaults etc. I would like to know what is causing my system to hang. Don -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 04/25/2015 04:31 PM, don fisher wrote:
On 04/25/2015 12:24 PM, Anton Aylward wrote:
man 3 syslog Thanks. I did not recognize the PID. Do you know how to start klogd?
Why do you want to run klog? Wickd is a user process. Are you sure you don't want a regular syslog? The fact that you are getting the messages says that the syslog system is working. You didn't say where you got the messages. That's why I said RTFM. Check, using 'ps' to see what you have running. That you are seeing those log entries at all makes me wonder. Where are they from? They don't seem to be in the class of "something said by the kernel" that would require klogd. Maybe your issue is configuring the syslogd. RTFM about syslog, such as the man page(s). The reason I mentioned the initial RTFM was that wickd uses syslogging. Wickd is an application. That why you saw the pid.. You can configure BOTH how wickd _issues_ syslog messages and how the syslog daemon listens for them and files them in the appropriate /var/log/ file but the appropriate config files under /etc. You want a formula there that filters the wickd daemon syslog messages to a specific destination. Again, RTFM. Nothing new here. that's how it works for so, so, so many applications.
I tried the usual:
sudo systemctl enable klogd.service sudo systemctl start klogd.service
but received an error message:
Failed to start klogd.service: Operation refused, unit klogd.service may be requested by dependency only.
Well have you looked at what's in the unit file? [Unit] Description=System Kernel Logging Service Requisite=syslog.service BindTo=syslog.service After=syslog.service RefuseManualStart=true
I know what a dependency is, just not what the message implies.
You can't start it manually, you can only start it as a dependency. Probably if you rebooted now, or restarted syslog.service, it would start. But I ask again, why do you want kernel logging for something that is an application, running in user space? This is a syslog issue not a klog issue. -- The state can't give you free speech, and the state can't take it away. You're born with it, like your eyes, like your ears. Freedom is something you assume, then you wait for someone to try to take it away. The degree to which you resist is the degree to which you are free... --Utah Phillips -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-04-26 02:19, Anton Aylward wrote:
You can't start it manually, you can only start it as a dependency. Probably if you rebooted now, or restarted syslog.service, it would start.
But I ask again, why do you want kernel logging for something that is an application, running in user space?
I don't think you need klogd to get kernel log entries in syslog (at least with systemd). I get them: <0.5> 2015-04-26 02:16:45 minas-tirith kernel - - - [ 1247.557051] REISERFS (device dm-0): Using r5 hash to sort names - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlU8MXwACgkQja8UbcUWM1zldQD9HP1U/9RA4uAvinuZMEPEyPgy FkPsdJMGMBGEfOccnQAA/Atz/XtyP8zatJGq3yKTfRR9wPYtnVu0NIROA8lAQrKz =kCjJ -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 04/25/2015 08:29 PM, Carlos E. R. wrote:
On 2015-04-26 02:19, Anton Aylward wrote:
You can't start it manually, you can only start it as a dependency. Probably if you rebooted now, or restarted syslog.service, it would start.
But I ask again, why do you want kernel logging for something that is an application, running in user space?
I don't think you need klogd to get kernel log entries in syslog (at least with systemd). I get them:
<0.5> 2015-04-26 02:16:45 minas-tirith kernel - - - [ 1247.557051] REISERFS (device dm-0): Using r5 hash to sort names
Now you mention it, so am I and I'm SURE I'm *NOT* running klogd. Don reall is rabbiting off on something spurious with this one. -- A: Yes. > Q: Are you sure? >> A: Because it reverses the logical flow of conversation. >>> Q: Why is top posting frowned upon? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 04/25/2015 04:31 PM, don fisher wrote:
I would like to know what is causing my system to hang.
If this were me I'd isolate the problem. Its sometimes called 'wolf-fencing'. http://dl.acm.org/citation.cfm?id=358695 <- Its well established :-) 1. Will the hard-wired connection set-up ONLY work? 2. Will the wifi connection ONLY work? 3. How are the relative priorities of the interfaces established? Is this to do with routing? Am I mistakenly setting up a bridge? 3. If I have the hard-wired working without the wifi, then try to start the wifi what happens 4. If if have the wifi working without the hard-wired, then try to start the hard-wired, what happens? 5. Is IP forwarding turned on? -- A: Yes. > Q: Are you sure? >> A: Because it reverses the logical flow of conversation. >>> Q: Why is top posting frowned upon? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 04/25/2015 05:29 PM, Anton Aylward wrote:
On 04/25/2015 04:31 PM, don fisher wrote:
I would like to know what is causing my system to hang.
If this were me I'd isolate the problem. Its sometimes called 'wolf-fencing'. http://dl.acm.org/citation.cfm?id=358695 <- Its well established :-)
1. Will the hard-wired connection set-up ONLY work?
2. Will the wifi connection ONLY work?
3. How are the relative priorities of the interfaces established? Is this to do with routing? Am I mistakenly setting up a bridge?
3. If I have the hard-wired working without the wifi, then try to start the wifi what happens
4. If if have the wifi working without the hard-wired, then try to start the hard-wired, what happens?
5. Is IP forwarding turned on?
Sorry, I thought you gave me more credit than that. My system hang is probably not wicked based. It usually occurs during boot. It may be my graphics system, but I do not know because I receive no information. I miss my rs232 port and the boot logs. I may have to get a usb to rs232 adapter. I am running an Alienware laptop, which has dual GeForce GTX 880M graphics GPUs with 16GB of video ram. This is not common hardware and I can not be sure how capable the nouveau driver is with it. But please do no flame me again. I will look into it myself. I was just hoping to get some data about the problem from the kernel itself to narrow the search. Don -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-04-26 03:33, don fisher wrote:
I miss my rs232 port and the boot logs. I may have to get a usb to rs232 adapter.
It is possible to tell the kernel to log to another machine via ethernet. I have done it, works right. Not as low level as a real serial port, but it makes do. Mostly :-) - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlU8RHoACgkQja8UbcUWM1yBSAD9HHP38Wt4WvckklTgtetbMATw ULoyiffAxjJ1wSbZ0o8A/iIMcLkuNHnA+SS2RbuMt/Mr+AjBbWDc2RpicQVuC2rt =pVEg -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 04/25/2015 09:50 PM, Carlos E. R. wrote:
On 2015-04-26 03:33, don fisher wrote:
I miss my rs232 port and the boot logs. I may have to get a usb to rs232 adapter.
It is possible to tell the kernel to log to another machine via ethernet. I have done it, works right. Not as low level as a real serial port, but it makes do. Mostly :-)
You'd be looking at ...../Documentation/networking/netconsole.txt However it says <quote> It can be used either built-in or as a module. As a built-in, netconsole initializes immediately after NIC cards and will bring up the specified interface as soon as possible. While this doesn't allow capture of early kernel panics, it does capture most of the boot process. </quote> So I don't know how useful it is. I've never needed to use it. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 04/25/2015 09:33 PM, don fisher wrote:
Sorry, I thought you gave me more credit than that. My system hang is probably not wicked based.
Sorry. You're not always a clear communicator. After all, this thread you started, this Subject: is about wicked (sic). Now you mention boot hand but you've been going on about networking conflict. Where does it clash?
It usually occurs during boot.
WHAT usually occurs during boot? The problem with wickd?
It may be my graphics system, but I do not know because I receive no information.
Yes you do, you just don't know where to look.
I miss my rs232 port and the boot logs. I may have to get a usb to rs232 adapter.
WTF? When I want to see the boot messages, when there is problems booting, long delays, or a hang., I edit the boot command line. At the grub2 prompt I type 'e' and get the editor for the line I've chose. It may be the normal one or it may be one of the "Advanced Options" such as a "recovery mode". Usually I skip the recovery mode unless I'm concerned with the graphics side of things. Usually I just remove the "splash=silent quiet". I then get a lot of streaming output, the kind of thing that back in the days of Linux 0.9x in a distant century past would, as you mention, require a serial hook-up. "Yes it used to be but we changed all that". hang? Well maybe not. A lot of the time I see a fsck rebuilding my disk from the logs. As mentioned, I use ResiserFS and its very good about logs and recovery, but this can take some time. I don't know about BtrFS. As for ext[234], YMMV. As I've also mentioned, I keep many past kernels and configurations. Its one way of doing debugging. There are many possible command line options for booting. Some have been mentioned in this forum in the past; they are documented both in the system docs - RTFM - as well as on-line - "go google". Try 'man 7 bootparam' for starters. You might be interested in things like 'panic=N' By default the kernel will not reboot after a panic, but this option will cause a kernel reboot after N seconds (if N is greater than zero). You can also set the kernel log level for those messages I mentioned. loglevel= level be paranoid, set that to 0 or 1 !!! I've also found that, since the messages can be confusing when the SMP kernel & systemd is running multiple fscks in parallel to disable multi-threading & SMP. KISS. -- A: Yes. > Q: Are you sure? >> A: Because it reverses the logical flow of conversation. >>> Q: Why is top posting frowned upon? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 04/25/2015 09:33 PM, don fisher wrote:
On 04/25/2015 05:29 PM, Anton Aylward wrote:
On 04/25/2015 04:31 PM, don fisher wrote:
I would like to know what is causing my system to hang.
[...]
My system hang is probably not wicked based. It usually occurs during boot. It may be my graphics system, but I do not know because I receive no information.
Actually you do get information, you just haven't told us. Its called context, and I'm well known for saying Context is Everything Where in the boot sequence is the context. It hangs WHERE? Do you get the BIOS self test? On my Dell machines I can set for short or detailed self test. Does your BIOS allow similar? Does it jet past the power on self test? The BIOS then loads the MBR. Where did you put the MBR? Is the MBR loading? IIR Felix has a lot to say on where you should put the MBR and why and what the side effects of putting it in the wrong place might be. Do you get into GRUB? If you can't is that because of a disk problem or because you have the MBR not leading to /boot? Does GRUB hang or do you get the menu? GRUB may just say "GRUB" or part of that word ... Then nothing. https://en.opensuse.org/GRUB https://help.ubuntu.com/community/Grub2/Troubleshooting if you get to the menu you should have the option of "recovery mode". Does that work? If you get the menu can you edit an entry at all? If, as you think, this has to do with your graphics adaptor, there a number of command line options to disable that, run in text mode, disable automatic or some features of how X talks to the graphics hardware. These are all documented. RTFM on kernel parameters. https://www.kernel.org/doc/Documentation/kernel-parameters.txt (leads to many more documents) One of the command line options is "init=" If you google you'll find examples of that for giving a shell rather than booting the system properly. Many other options for working with, for example ACPI, interrupts etc. One I mentioned was turning SMP off so that you're not confused by multiple threads. Systemd will ytry running many fscks in parallel and that cab be very confusing. As for the serial adapter idea, or its network equivalent, if you really are in dire straits and the problem is before GRUB gets the initrd to run and the kernel to get as far as reading the command line and setting up that channel, then its not going to be a lot of use. That's my experience. I've never set up such a channel, always managed to do my boot problem debugging with the above sequence or by using the rescue disk. Most of my boot problems have involved lvmetad, and while the kernel messages go quickly when I take out "quiet" that has never mattered. Its ALWAYS been what happens last before the hang/crash occurs and that stays on screen :-) Grub can let you test the disk, search for something to load. Google for that. of course there is an implicit assumption in all this: that you have a second machine connected to the 'net so you can google, search the opensuse wiki, google in general, read the documentation. This kind of debugging is mix of being holistic, that is comprehensive, considering all those possibilities, and of being reductionist, listing all the possible 'failure modes' and their paths and effects. I admit I'm more explicit in this, enumerating the check list. I was taught to keep a log not only of what happens but also of my decision-tree, why I chose to investigate a path, how I interpreted the evidence, so I can back-track and see if my assumptions were wrong. The only problem with this is that, as the FBI, police deparments and lawyers have found, is that searching all those notes can become difficult. That's why we have wikis, not the MediaWiki ones but ones that do automatic linking. Sadly they don't help with records on "dead trees". Keeping records addresses a human shortcoming. We think we can hold things in our head, but the reality is that we can only hold a few in short term memory and to hold them in long term memory means a more formal mechanism. That long term memory might be the notebook or wiki. -- A: Yes. > Q: Are you sure? >> A: Because it reverses the logical flow of conversation. >>> Q: Why is top posting frowned upon? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
participants (3)
-
Anton Aylward -
Carlos E. R. -
don fisher