your opinion on Xymon HA (or dual server config). 🔗 link

Hi:
The new site administrator has been happy with the simplicity and usability of Xymon.
Then have asked for more changes over our actual config.
Currently I have one (kind of resource limited) Sparc V210 as xymon server. But have find out then sometimes the apache or xymond process coredump by some mistakes or changes. Sometimes we notice when this happen so just restart the process or server. But when the problem arises and we don't notice the problem; after 8 hours there is a lot of rrd data that can't be showed. (We can't send sms or emails out of the site domain).

So my boss have asked to make an another xymon server; to workout changes on one and keep the other with the older config till the new changes are "stable".
I have read the HA section but I'm unsure that can justify a cluster. So my question is if I have two xymon servers with the same config and tell the clients that report to both of them; can this scenario accomplish the task ?
I think that my boss is requesting a kind of pre-production/ production config. So I though two same config servers fit the scenario or must use the cluster ??

Any input is welcome!

Thanks a lot!


-- 

CONFIDENCIALIDAD - La información contenida en este mensaje es confidencial y se dirige únicamente a su destinatario. Si usted lector de este mensaje no es ese destinatario, la diseminación, distribución o copia del mismo o sus adjuntos (de existir) se encuentran prohibidos. Si lo ha recibido por error, por favor notifique de manera inmediata por correo y destruya las copias de su correo. 
CONFIDENTIALITY STATEMENT - The information contained in this message is confidential and intended only for the addressee. If the reader of this message is not the intended recipient you are notified that any dissemination, distribution or copy of this message and attachments (if any) is strictly prohibited. If you have received this in error, please immediately notify us by reply email, destroy all copies and remove from all media.

Randall

I have much the same setup as you have suggested, although they are both
considered production, rather than one being "pre-prod".  Essentially, both
Xymon servers operate completely independently of each other, so if either
server goes down, the other server continues to receive all of the client
updates and do all of the probes.

I have all of the configuration committed to a subversion server.  So I can
make a change on one Xymon server, test it, commit it and then "svn update"
on the other server to pull in the change.  All configuration files between
the two servers are identical, with the exception of a few variables in
xymonserver.cfg.  So what I do is create myself a file
xymonserver-local.cfg, and include it near the top of xymonserver.cfg.  In
the "-local" version of the file, I define XYMONSERVERHOSTNAME,
XYMONSERVERIP and XYMONSERVERS to be the name and address of that Xymon
server.  You should remove any of these definitions from the main
xymonserver.cfg and add the following line near the top:

include /etc/xymon/xymonserver-local.cfg

Both of my servers have every client host listed in the replicated
hosts.cfg.  So each Xymon server independently pings and tests each host in
its network checks.  In each of the clients hosts, I list both server IPs
in XYMONSERVERS, so that both servers get client messages from the client,
and status messages from any ext scripts they might happen to run.

One shortcoming with what I've described is that disabling or enabling a
server will only apply to the Xymon server where that update takes place.
But there's a simple way to have these changes propagate to both servers.
First, create another file called xymonserver-enadis.cfg, containing the
following:

include /etc/xymon/xymonserver.cfg
XYMSERVERS="10.1.1.1 10.2.2.2"

In this file, list the IPs of both Xymon servers in XYMSERVERS.  Now,
update cgioptions.cfg and set:

XYMONENV_ENADIS=/usr/lib/xymon/server/etc/xymonserver-enadis.cfg

Now, when you disable a host, this will run xymonserver-enadis.cfg but have
it pull its config from the special "-enadis" file, complete with the
XYMSERVERS parameter listing both IP addresses.  So disabling a host on any
one Xymon server causes the disable message to be sent to both servers.

There's one thing to watch out for with this setup, is when you're testing
changes to client-local.cfg made to only one of the Xymon servers.  When
each client sends it client data, it is also given its section of
client-local.cfg to process.  If the client sends client data to two Xymon
servers, it's the client-local.cfg section from the second Xymon server
that will be used, and any "test" changes to that file on the first server
will be ignored.

Also, check out the WikiBook "HA" section for some other options:
https://en.wikibooks.org/wiki/System_Monitoring_with_Xymon/Administration_Guide#Hobbit_in_HA

Cheers
Jeremy

On Wed, Dec 16, 2015 at 9:11 AM Randall Badilla Castro <

user-911f8b61a6eb@xymon.invalid> wrote:

I have encountered similar issues in the past and opted for the following script that I run via cron every 10 minutes.

It looks at the number of Xymon processes and if there are too few it attempts a restart and alerts me of the restart.  It is very simple, it is not perfect, but it does work (for me).

#!/bin/sh

PROC=`ps -eaf |egrep 'xymond|hobbitd'|egrep -v 'grep|ps -eaf'|wc -l`
PROC2=`ps -eaf |egrep 'xymond|hobbitd'|egrep -v 'grep|ps -eaf'`
HOSTNAME=`hostname`

echo "------------- `date` --------------" >/home/xymon/logs/am_I_running.log

ERROR=0

if [ "$PROC" -lt 1 ];then
   sleep 5
   PROC=`ps -eaf |egrep 'xymond|hobbitd'|egrep -v 'grep|ps -eaf'|wc -l`
   if [ "$PROC" -lt 1 ];then
      ERROR=1
   fi
fi

if [ "$PROC2" = "" ];then
   sleep 5
   PROC2=`ps -eaf |egrep 'xymond|hobbitd'|egrep -v 'grep|ps -eaf'`
   if [ "$PROC2" = "" ];then
      ERROR=1
   fi
fi

if [ "$ERROR" -eq 1 ];then
   echo "Number of processes: $PROC" >/home/xymon/logs/am_I_running.log
   echo $PROC2 >>/home/xymon/logs/am_I_running.log
   /home/xymon/server/xymon.sh start
   echo "** Restart attempt has been made." >>/home/xymon/logs/am_I_running.log
   mailx -s "$HOSTNAME Xymon may not be running - attempting restart" <youremail>@<yourdomain> </home/xymon/logs/am_I_running.log
   mailx -s "$HOSTNAME Xymon may not be running - attempting restart" <pagernumber>@<textsms> </home/xymon/logs/am_I_running.log
else
   echo "Number of processes: $PROC" >>/home/xymon/logs/am_I_running.log
fi

Thanks,
John
Upcoming PTO: 12/8, 12/15 & 12/16
John Rothlisberger
IT Strategy, Infrastructure & Security - Technology Growth Platform
TGP for Business Process Outsourcing
Accenture
XXX.XXX.XXXX office

-----Original Message-----
From: Xymon [mailto:xymon-bounces at xymon.com] On Behalf Of Randall Badilla Castro
Sent: Tuesday, December 15, 2015 4:12 PM
To: xymon at xymon.com
Subject: [Xymon] your opinion on Xymon HA (or dual server config).

Hi:
The new site administrator has been happy with the simplicity and usability of Xymon.
Then have asked for more changes over our actual config.
Currently I have one (kind of resource limited) Sparc V210 as xymon server. But have find out then sometimes the apache or xymond process coredump by some mistakes or changes. Sometimes we notice when this happen so just restart the process or server. But when the problem arises and we don't notice the problem; after 8 hours there is a lot of rrd data that can't be showed. (We can't send sms or emails out of the site domain).

So my boss have asked to make an another xymon server; to workout changes on one and keep the other with the older config till the new changes are "stable".
I have read the HA section but I'm unsure that can justify a cluster. So my question is if I have two xymon servers with the same config and tell the clients that report to both of them; can this scenario accomplish the task ?
I think that my boss is requesting a kind of pre-production/ production config. So I though two same config servers fit the scenario or must use the cluster ??

Any input is welcome!

Thanks a lot!


--

CONFIDENCIALIDAD - La información contenida en este mensaje es confidencial y se dirige únicamente a su destinatario. Si usted lector de este mensaje no es ese destinatario, la diseminación, distribución o copia del mismo o sus adjuntos (de existir) se encuentran prohibidos. Si lo ha recibido por error, por favor notifique de manera inmediata por correo y destruya las copias de su correo.

CONFIDENTIALITY STATEMENT - The information contained in this message is confidential and intended only for the addressee. If the reader of this message is not the intended recipient you are notified that any dissemination, distribution or copy of this message and attachments (if
any) is strictly prohibited. If you have received this in error, please immediately notify us by reply email, destroy all copies and remove from all media.

This message is for the designated recipient only and may contain privileged, proprietary, or otherwise confidential information. If you have received it in error, please notify the sender immediately and delete the original. Any other use of the e-mail by you is prohibited. Where allowed by local law, electronic communications with Accenture and its affiliates, including e-mail and instant messaging (including content), may be scanned by our systems for the purposes of information security and assessment of internal compliance with Accenture policy.

www.accenture.com