June 2016 - Xymon - Xymon mailman web

Find server sending data to xymon?
by Russell.Blumenthal＠stonybrookmedicine.edu 07 Jun '16

07 Jun '16

Hi, Is there a way to find out what server is sending data to xymon? I have an entry where the status has been flapping. Looking at it more, it looks like 2 servers are sending data using the same hostname for the xymon config. I disabled the service on the known server (good one) to stop the flapping, but I am not sure how to find out who is sending the other data. Any ideas? Thanks This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by e-mail and destroy all copies of the original.

1 1

Recall: Find server sending data to xymon?
by Russell.Blumenthal＠stonybrookmedicine.edu 07 Jun '16

07 Jun '16

Blumenthal, Russell would like to recall the message, "Find server sending data to xymon?". This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by e-mail and destroy all copies of the original.

1 0

Support for TLS v1.1 and 1.2?
by ian.burrowes＠gmail.com 06 Jun '16

06 Jun '16

Hi David, This might not help at all, but when I had problems with xymon HTTPS and SSL checks on my Win2012 R2 ADFS server recently, it was actually the target website requiring Server Name Indication (SNI) support, rather than a TLS 1.1 / 1.2 issue. I was able to fix the problem by adding 'sni' on the end of the host entry in hosts.cfg. Please disregard this advice if not relevant, but perhaps it will help others :) Kind Regards, Ian On 7 June 2016 at 08:31, Gore, David W (David) via Xymon <xymon at xymon.com> wrote: > _______________________________________________ > Xymon mailing list > Xymon at xymon.com > http://lists.xymon.com/mailman/listinfo/xymon > > > ---------- Forwarded message ---------- > From: "Gore, David W (David)" <david.gore at verizon.com> > To: "xymon at xymon.com" <xymon at xymon.com> > Cc: > Date: Mon, 6 Jun 2016 18:20:12 -0400 > Subject: [Xymon] Support for TLS v1.1 and 1.2? > > Mark Felder, > > > > Mentioned last year around April 17th, 2015 where Xymon support for TLS > v1.1 and v1.2 may be lacking. Perhaps the issue is more my naiveté but > does anyone know how I can get the sslcert and http tests to work correctly > with Apache and Xymon. > > > > [image: red] https://xymon1.domain.com/ - SSL error > > > > The sslcert test goes purple. > > > > Os: Red Hat Enterprise Linux Server release 7.2 (Maipo) > > Openssl: OpenSSL 1.0.1e-fips 11 Feb 2013 > > Xymon: 4.3.26 > > > > > > David W Gore > > > >

1 0

How to override testip?
by feld＠feld.me 03 Jun '16

03 Jun '16

How do you override testip for a single host? Looking at the code there's a flag for xymonnet called "dodns" that's not documented in hosts.cfg man page. It feels to me like the xymonnet code should be slightly altered to allow you to specify "dodns" on a per-host basis to override a "testip" setting that may be default. -- Mark Felder feld at feld.me

2 2

Xymon Client Short Hostname (was: Re: xymon for AIX)
by wonderfoo2＠gmail.com 02 Jun '16

02 Jun '16

There were two issues that seemed to have resolved - related to the AIX clients. Both were resolved through the tips provided along these threads. One was opening up the Centos server firewall ports AND the suggestion provided by Ryan to cure the state with client web pages not getting updates - shown as 'Ghost Clients' under 'Reports' (no updates until the short hostnames were populated in host cfg on the xymon server). All of our *NIX hosts were of FQDN hostnames in this environment. Hope that clarify the 'mystery' surrounding AIX client for xymon on this thread. Thank you ALL! On Tue, May 31, 2016 at 4:27 PM, Ryan Novosielski <novosirj at rutgers.edu> wrote: > You are most welcome. > > So here’s some followup: many of us have had this problem before, and it > seems OS-specific. Is there any way to get this more right by default? I > see that many/most of my entries in hosts.cfg have CLIENT:<hostname> > specified. My server is Solaris 10, my clients are a mixture of things. > > > On May 31, 2016, at 5:10 PM, L Foo <wonderfoo2 at gmail.com> wrote: > > > > Thanks for ALL your valuable feedback! > > > > The hint from Ryan seemed to have cured the client page reporting > issues. As it turned out, the AIX clients in questioned showed up in 'Ghost > Clients' list on in short hostname as opposed to it's real FQDN hostname. > I've since updated the xymon server side configuration's files for the said > AIX clients from FQDN hostname to short hostnames, the client pages were > getting proper updates now. Thank you Ryan!!! > > > > On Thursday, May 26, 2016 12:05 AM, Jeremy Laidman < > jlaidman at rebel-it.com.au> wrote: > > > > > > On Thu, May 26, 2016 at 4:44 AM L Foo <wonderfoo2 at gmail.com> wrote: > > The netcat to port 1984 seems to have went through/connected ok based on > below output: > > > > # nc -v 172.31.2.131 1984 > > Ncat: Version 6.40 ( http://nmap.org/ncat ) > > Ncat: Connected to 172.31.2.131:1984. > > > > You ran this on the AIX host? If this works, then telnet should have > also worked, and so should the xymon client. > > > > Can you please try the following (on the AIX host): > > > > # echo "ping" | nc 127.31.2.131 1984 > > > > That should return the xymon version details of the Xymon server. > > > > Then try this: > > > > # /xymon/bin/xymon 172.31.2.131 ping > > > > You should see the same thing. Or maybe a timeout > > > > If this doesn't work, then can you try this: > > > > # truss -f /xymon/bin/xymon 172.31.2.131 ping > > > > I'm particularly interested in the output around socket() and connect() > calls, and what responses codes are given. > > > > I'm wondering if you have some sort of kernel-based binary restrictions > in place, such as Trusted AIX, TCB or Trusted Execution. From what I can > tell, these tools allow one to lock down a server so that no unauthorized > binaries can execute, or they might be able to execute but cannot perform > certain functions. > > > > An alternative to all of this is to use my xymon-rclient script, > available on xymonton.org, or http://tools.rebel-it.com.au/xymon-rclient. > Essentially it allows for an clientless client by pushing the Xymon client > script to the target from the Xymon server (typically over ssh) and grabs > the client data directly. In this way, it doesn't rely on a xymon client > binary. There are some limitations, which is why it's better to try to get > your client working, but it may be easier to get something going this way. > > > > Cheers > > Jeremy > > > > > > > > -- > ____ > || \\UTGERS, |---------------------------*O*--------------------------- > ||_// the State | Ryan Novosielski - novosirj at rutgers.edu > || \\ University | Sr. Technologist - 973/972.0922 (2x0922) ~*~ RBHS > Campus > || \\ of NJ | Office of Advanced Research Computing - MSB C630, Newark > `' > >

1 0

question
by jlaidman＠rebel-it.com.au 02 Jun '16

02 Jun '16

On Mon, May 30, 2016 at 10:52 PM <m.pingel at ndr.de> wrote: > is it possible to configure a maximum level of memory from a process? > Let me rephrase to make sure I understand. You want to monitor the memory usage of a process (by name, presumably) and if it exceeds a predefined threshold, you would like it to change colour? This is not really possible out-of-the-box. However, there are ways to do this. The xymon client data message contains the "ps" section to list all running processes. Generally, this includes memory usage for each process. So the Xymon server has the data for it to use. Example: https://www.xymon.org/xymon-cgi/svcstatus.sh?CLIENT=claudio.hswn.dk&section… Note that there are several columns you might be interested in here. An absolute MB of memory is shown in the resident size (RZ) and virtual size (VZ), but also the percentage of memory (%MEM). So, step one is to decide what you're interested in, perhaps based on what problem you're trying to detect. If you wanted to alert on client data, you could write a server-side task to query every device and do the thresholding. You could either create a new test such as "procmem", or you could modify the state of the "proc" test. Here's a scriptlet that would probably (untested) do the latter: #!/bin/sh PROCMATCH="/usr/sbin/ntpd" # should fetch per-host with hostinfo or xymongrep $XYMON $XYMSRV "xymondboard test=proc fields=hostname,client" | grep ",Y$" | while read LINE; do HN=`IFS="|"; set - $LINE; echo $1` PS=`$XYMON $XYMSRV "clientlog $HN section=ps" | grep "$PROCMATCH"` # parse PS for the appripriate column and generate a "modify" message for the "prox" status done Sorry, I don't have access to a Xymon server to test this with just now. J

2 1

Xymon custom colour (wallboard) page
by spah＠syntec.co.uk 02 Jun '16

02 Jun '16

Hi, Is there an easier way to create a xymon page with a custom colour listing (while keeping the existing nongreen and red.html pages untouched) than adding the following to xymongen.c line 658 and recompiling? nongreencolors = (red,purple); nongreen_color = do_nongreen_page(nssidebarfilename, PAGE_NONGREEN, "wallboard"); * I'm not sure if this is the correct syntax for nongreencolors. I would also like to remove the info and trends column from this page to make a wallboard page, but if I used --nongreen-ignorecolumns=info,trends it would also affect the existing nongreen page that I don't want to be changed. Or is there a better alternative for creating an html wallboard type page? Kind regards, SebA

1 0

Xymon all red page minor bug question and patch
by spah＠syntec.co.uk 02 Jun '16

02 Jun '16

Hi, I found the Xymon all red page (red.html) added in 4.3.11 (r7180) with a bit of Googling, but (a) it is not in xymonmenu.cfg so it is not visible by default (trvial patch to config attached to fix this), (b) (the minor bug) it displays 'Current non-green Systems' at the top of the page instead of e.g. 'Current red Systems', and (c) is there any way it can display the most recent events, like the actual nongreen.html page does? I did look at the diffs from the commit that added this feature, but I couldn't see anything there to customize the header of the page (or any reason why it would be suppressing the most recent events view). Kind regards, Sebastian Auriol

1 0

acknowledge alert
by Jurgen.Claessens＠cegeka.be 02 Jun '16

02 Jun '16

Hello all, Currently we are still running xymon 4.3.17 on our servers, but are looking to update them to the latest version. We acknowledge alerts thru a custom webpage layout that sends it to our 2 servers. In 4.3.17 this worked like a charm, but on the newer versions this does no longer work. I know from 4.3.18 things have been made more secure in part of the acknowledge.sh and stuff, but is there still a way to acknowledge thru http? Best regards, Jurgen Claessens This e-mail and all files transmitted as attachment(s) thereto are confidential and solely intended for the individual to whom or the organization to which they are addressed. If you received this e-mail by mistake, please notify Cegeka's Service Desk at cegeka.support at cegeka.be or call +32 (0)11 240 363. We thank you in advance. Cegeka hereby confirms that this message has been swept by Sophos for the presence of viruses.

1 0

Xymon Client Short Hostname
by stef.coene＠docum.org 01 Jun '16

01 Jun '16

On 05/31/2016 11:27 PM, Ryan Novosielski wrote: > You are most welcome. > > So here’s some followup: many of us have had this problem before, and it seems OS-specific. Is there any way to get this more right by default? I see that many/most of my entries in hosts.cfg have CLIENT:<hostname> specified. My server is Solaris 10, my clients are a mixture of things. In theory, a hostname contains no domain suffix. So you should never have a FQDN in your hosts.cfg Stef

5 4