February 2014 - Xymon - Xymon mailman web

FreeBSD [ifstat] processing
by jlaidman＠rebel-it.com.au 20 Feb '14

20 Feb '14

Hiya I have some FreeBSD systems that run IPv6, and the numbers that Xymon are graphing are not quite right. I think I've pinned down the cause as being the ifstat matching regexp, not being able to handle either multiple bound IP addresses, or IPv6 addresses. My system has output from "netstat -i -b -n" something like this: Name Mtu Network Address Ipkts Ierrs Ibytes Opkts Oerrs Obytes Coll em0 1500 <Link#1> aa:bb:cc:dd:ee:ff 855109239 246 119882177806 4445784528 0 1678631985244 0 em0 1500 fe80:1::666:6 fe80:1::666:66ff: 5652 - 403008 8487 - 565928 - em0 1500 192.168.1.0 192.168.1.66 3669836 - 327402522 4441615751 - 1612495913591 - em0 1500 6006:6000:0:6 6006:6000:0:6006: 186828 - 13051632 33119395 - 3319372758 - <snipped> lo0 16384 <Link#3> 2420387 0 296603593 2420667 0 296585785 0 lo0 16384 ::1/128 ::1 193448 - 17014491 217323 - 22687570 - lo0 16384 fe80:6::1/64 fe80:6::1 0 - 0 0 - 0 - lo0 16384 127.0.0.0/8 127.0.0.1 2227042 - 279599853 2226396 - 279550576 - lo1 16384 <Link#4> 398523 0 55461313 398523 0 55461150 0 lo1 16384 192.168.2.0 192.168.2.66 317552009 - 63772241684 383671 - 50695917 - lo1 16384 6006:6000:600 6006:6000:600::66 29688469 - 10867392950 4 - 2982 - <snipped> The Xymon client explicitly excludes the "link" lines, and all of the "lo" lines. Makes sense, except that it doesn't work for me. It seems that the Xymon server-side parsing is only picking up the line with the IPv4 address (192.168.1.66) and ignores all the others for the interface. Also, I can't see where Xymon handles multiple lines for the same interface, and I don't think it does. And finally, this server has loopback interfaces lo1, lo2 and so on, with alternative IP addresses for this server. Most of the traffic that arrives at this server is addressed to the loopback interfaces. So even if Xymon could parse these lines, multiple of them, the client script excludes all of the loopback interfaces. What's the best way to solve this? The lo[1-n] NICs could be added back in. But I don't know that this is very useful. The intent of the [ifstat] data is to record and display bytes that come in and out of physical NICs that communicate with the outside world, but my extra loopback interfaces' traffic isn't being counted. For this reason, it seems to make sense that the "link" lines are probably the best for this, as I would think that they would show packets contained only in physical frames (ethernet or some other medium). However, I think the "link" lines also show things like ARP and other layer-2 traffic that might not be as useful. Has anyone else looked into this, or done monitoring of FreeBSD interfaces? Is there a better command to run on FreeBSD for this? I'd really like to get this fixed up. Cheers Jeremy

2 3

FILES in analysis.cfg
by Nick＠Pettefar.com 19 Feb '14

19 Feb '14

Is there also a way to have the red with a wildcard? I am thinking specifically about /var/core/core* files which are very important to know if they exist but have undetermined names. We also need the alarm to stay on until the core* files are removed. Regards, Nick Pettefar > Date: Tue, 18 Feb 2014 22:54:30 +0100 > From: Henrik St?rner <henrik at hswn.dk> > Subject: Re: [Xymon] FILES in analysis.cfg >> So how do I get the files test go RED on a missing file? > > OK, I had to do some testing with this. > > The problem is that when the file does not exist, then the `ls -1 ...` > will not produce any output. Without any output, Xymon cannot tell that > there is a file-check happening. > > If you have an explicit filename in client-local.cfg, then the FILE > check in analysis.cfg works as expected, because the xymon client will > report that the specific file is missing. > > So: > > client-local.cfg: > [myhost] > file:/var/tmp/sys-heartbeat-myhost > > analysis.cfg > HOST=myhost > FILE %^/var/tmp/sys-heartbeat-.* MTIME<3700 > > This will go red if the file is missing, or it exists and is older than > 3700 seconds. > > > Regards, > Henrik Call Send SMS Add to Skype You'll need Skype CreditFree via Skype

2 1

Cannot get custom script column back.
by ghewit01＠harris.com 19 Feb '14

19 Feb '14

Hi, I'm having problems with how xymon is reporting a status and wanted to give you a step-by-step detailed description of what I did to get in the state that it is in so that it will hopefully describe what to do to fix this anomally. I set up a custom script to monitor the status of process, more than just monitoring to see if the daemon is up or down, etc, etc. In testing the script I used the example of the client-side script on my client. I created the script on the clients and everything showed up on the xymon webpage and worked perfectly. I then realized that I wanted to rename the script on the client to something better describing it. So I proceeded to rename the script and update the clientlaunch.cfg in etc on each client to reflect the new name. After a short period the column on the xymon website showing the status of the original name of the script went purple because that script name was not reporting anymore. I removed this purple column by running xymon 127.0.0.1 "drop <HOSTNAME> <columname>". The problem is I can't get the column to show back up again with the status from the newly named script that is currently configured. The names of the columns were the same, just the script name changed. I suspect that is why it is behaving this way but don't know what to do to fix this state that it's in. Any help would be greatly appreciated. Lenny

3 2

XYMON Proxy Issue
by gbegin＠csc.com 19 Feb '14

19 Feb '14

Hello I'm using XYMON 4.3.12 under Solaris 10.5 Since I put a proxy between my XYMON server and ~1000 agents, I have such pbl: - Many test are in pruple state, especially BBWin agents. - Even if the test is green, nothing is displayed in RRD graph. - I find timeout messages in logs (agents and server and xymonnet on the proxy): 2014-02-17 15:58:32 -> Recipient '10.195.243.205', timeout 15 2014-02-17 15:58:32 -> 1st line: 'combo' 2014-02-17 16:02:53 Whoops ! Failed to send message (timeout) - The xymonproxy log always complaining with a message that I don't understand: 2014-02-17 12:59:30 xymonproxy version 4.3.12 starting 2014-02-17 12:59:30 Listening on 0.0.0.0:1984 2014-02-17 12:59:30 Sending to Xymon server(s) 10.195.241.64:1984 2014-02-17 12:59:30 select() failed: Invalid argument 2014-02-17 12:59:30 select() failed: Invalid argument 2014-02-17 12:59:30 select() failed: Invalid argument - Many connections in TIME_WAIT from the proxy to the server, sometime the status is SYN_SENT - Tests coming from a second xymonproxy with fewer BBWin agents ( ~70) are working fine. I tried to play with timeout parameters on servers with no succès. Any idea ? Cordialement, Regards,Mit freundlichen Grüßen, Gautier BEGIN

1 3

clientlaunch.cfg question
by bakgat8＠hotmail.com 18 Feb '14

18 Feb '14

The command is defined by the CMD keyword. This is the full command including any options you want to use for this task. This is required for all tasks. When is says including any options you want to use for this task. what does it mean? I was hoping I can pass in some arguments to my script, but it does not work. So what would be an example of an option for this task? Befuddled. Bakkies

2 2

Step by step xymon instalation on CentOS-6.5-x86_64
by mate33＠ymail.com 17 Feb '14

17 Feb '14

Hello xymon community! I'm trying to insatall xymon server on my virtual centos box, but I can't get it to work. I would be extremely thankful if some has any kind of step by step instructions, because I'm not very skilled with linux (i'm still leraning). I think xymon is a great tool and I would really like to get it working. Tnx in advance. Best regards, Mate

2 1

opentsdb
by sean.clark＠twcable.com 17 Feb '14

17 Feb '14

JC Cleaver? I though you had mentioned this? From: <Clark>, Sean Clark <sean.clark at twcable.com<mailto:sean.clark at twcable.com>> Date: Thursday, February 13, 2014 at 3:36 PM To: Xymon Email List <xymon at xymon.com<mailto:xymon at xymon.com>> Subject: [Xymon] opentsdb I though I remembered seeing on here that someone had used —extra-script on xymon_rrd to send data to opentsdb… If that’s the case, do you have any points/settings for setting it up? I just setup an opentsdb instance, and wanted to move data into it. Thanks ________________________________ This E-mail and any of its attachments may contain Time Warner Cable proprietary information, which is privileged, confidential, or subject to copyright belonging to Time Warner Cable. This E-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this E-mail, you are hereby notified that any dissemination, distribution, copying, or action taken in relation to the contents of and attachments to this E-mail is strictly prohibited and may be unlawful. If you have received this E-mail in error, please notify the sender immediately and permanently delete the original and any copy of this E-mail and any printout.

1 0

4.3.16 make depends on gnu find/xargs
by john.thurston＠alaska.gov 15 Feb '14

15 Feb '14

There are several features of 4.3.16 I wanted (I'm currently running 4.3.12) but have been tripped up during the 'make install' process. ./build/Makefile.rules is using 'find -0' and 'xargs -0'. Neither option is familiar to me, but some reading of man pages indicates that these are options available on gnufind and gnuxargs (which I don't have on my Solaris systems). From http://www.gnu.org/software/findutils/manual/html_mono/find.html > using the find action ‘-print0’ and the xargs option ‘-0’ to avoid misinterpreting files whose names contain spaces It appears twice: > # find . -print |xargs grep "\-print0" > ./build/Makefile.rules: find $(INSTALLROOT)$(INSTALLWWWDIR) -type f -print0 | xargs -0 chmod 644 > ./build/Makefile.rules: find $(INSTALLROOT)$(MANROOT) -type f -print0 | xargs -0 chmod 644 Comparing with the Makefile.rules from 4.3.12 shows these lines are new. It looks to me like their purpose is to chmod everything in ~/server/www (the default location for the www files). I've looked in ~/server/www and can't find any files with spaces, so I'm going to modify these lines in MAkefile.rules (remove the -print0 and -0) and see how it goes. Are we expecting there to be files containing spaces in ~/server/www, or is -print0/-0 just considered best practice and the safe way to do this? <offtopic> Yes, I understand that everyone else in the world has all of the gnu tools installed by default and that my Solaris systems are stone-age relics. Sorry. IMHO, files don't have spaces in their names, and servers don't have frame buffers. I break the knuckles of anyone I find putting spaces in file names on servers. </offtopic> -- Do things because you should, not just because you can. John Thurston 907-465-8591 John.Thurston at alaska.gov Enterprise Technology Services Department of Administration State of Alaska

6 6

ssl errors since updating from 4.3.12 to 4.3.16 with openssl 1.0.1
by john.thurston＠alaska.gov 14 Feb '14

14 Feb '14

Since updating one of my xymon servers from 4.3.12 to 4.3.16, I have been seeing a lot of flapping on https tests to several tomcat servers. These failures are accompanied by lines in xymonnet.log of the form: > Unspecified SSL error in SSL_connect to 8443/tcp on host 10.203.10.42: error:00000000:lib(0):func(0):reason(0) I've just noticed that: 4.3.12 was built with openssl 0.9.8w 4.3.16 was built with openssl 1.0.1e These hosts were reporting well on the 4.3.12 instance of this Xymon server. They are still reporting well on my other (4.3.12) server. This appears, to me, to be a failure at establishing an SSL connection so I strongly suspect a difference in openssl behavior is the cause. The flapping does not appear to be occurring with https checks against my apache web servers, nor against my IIS web servers. Only against instances of tomcat. I'm going to go explore the differences between 0.9.8 and 1.0.1. My hypothesis is the cause is a change in cipher lists with 1.0.1. Has anyone else already seen these failures and found the cause? -- Do things because you should, not just because you can. John Thurston 907-465-8591 John.Thurston at alaska.gov Enterprise Technology Services Department of Administration State of Alaska

2 1

Cannot get custom script column back.
by ghewit01＠harris.com 14 Feb '14

14 Feb '14

Hi, I'm having problems with how xymon is reporting a status and wanted to give you a step-by-step detailed description of what I did to get in the state that it is in so that it will hopefully describe what to do to fix this anomally. I set up a custom script to monitor the status of process, more than just monitoring to see if the daemon is up or down, etc, etc. In testing the script I used the example of the client-side script on my client. I created the script on the clients and everything showed up on the xymon webpage and worked perfectly. I then realized that I wanted to rename the script on the client to something better describing it. So I proceeded to rename the script and update the clientlaunch.cfg in etc on each client to reflect the new name. After a short period the column on the xymon website showing the status of the original name of the script went purple because that script name was not reporting anymore. I removed this purple column by running xymon 127.0.0.1 "drop <HOSTNAME> <columname>". The problem is I can't get the column to show back up again with the status from the newly named script that is currently configured. The names of the columns were the same, just the script name changed. I suspect that is why it is behaving this way but don't know what to do to fix this state that it's in. Any help would be greatly appreciated. Lenny

2 3