Zenoss ZenTech Community

Is it possible to change the variable SERVICED_MASTER_IP (Default is the docker ip address: 172.17.42.1) on serviced config file with Physical interface IP Address (i.e eth0)

This variable define the serviced instance, but the ip address example for the serviced config file is alwalys Docker IP Address.

I have a possible conflict with two zenoss servers. Both are configured for defultt as Master and Agent simultaneusly (the serviced default config) with the IP 172.17.42.1 on SERVICED_MASTER_IP variable (OK because both pointing to its internal dokcer0 interface). The error is when two servers starts at the same time, one server adquire the Master role, and controls the another server (control center has the same information, you stop Zenoss corre process and these are stopped on both control center web pages). So I beliebe the proclaimed MASTER overwrite the data volume (local and remote - /opt/serviced/var/volumes - Is this possible i.e via the NFS export published by serviced that MASTER write both repositories.

So after this error, I beliebe the problem is the communication of the both servers, then I stop one server and reboot another for way to isolate the problem. But when you reboot server, now it has the data (control center names, Host, resource pools) than The proclamed server.

On my schema, both servers are independent servers, but on the same VLAN / subnet with this data:

--ZENOSS1---
--Phy Ip add: 172.16.9.1
docker0: 172.17.42.1
serviced config:
SERVICED_AGENT=1
SERVICED_MASTER=1
SERVICED_MASTER_IP=172.17.42.1

--ZENOSS2
--Phy Ip add: 172.16.9.2
-- docker0: 172.17.42.1
serviced config:
-- SERVICED_AGENT=1
-- SERVICED_MASTER=1
-- SERVICED_MASTER_IP=172.17.42.1

Subject:	It sounds like you're trying
Author:	Andrew Kirch
Posted:	2015-11-05 12:17

It sounds like you're trying to achieve high availability (HA), you might want to read the guide: https://support.zenoss.com/hc/en-us/articles/204666179-How-to-Configure-...

Andrew Kirch

akirch@gvit.com

Need Zenoss support, consulting or custom development Look no further. Email or PM me!

Ready for Distributed Topology (collectors) for Zenoss 5 Coming May 1st from GoVanguard

Subject:	Thanks for the information It
Author:	[Not Specified]
Posted:	2015-11-06 05:38

Thanks for the information It is a great document. Currently I'm trying to isolate two zenoss v5 servers, both are on the same VLAN but each on a remote site. The option HA Active - Passive is not necessary on this scenario.

My Active/Active scenario worked fine on same VLAN, until both servers booting at the same time. When both start, the control center search a new master (as both are on the same VLAN) if the one other server answers as Master then the problem described on the first post occurs.

So, one valid solution I had thought is an Active Active geographical scenario. After to deny the TCP /UDP ports required between control center daemons or to find a solution with SERVICED_MASTER_IP variables.

If it's possible, the next step is implement my CMDB syncronization between both servers (with my personal python scripts here - https://github.com/jmorcar/Zenoss.ZenEventClassMapping)

It's because I need two independent server collectors with their own web interface and own: SNMP/Ping polling (zenmodeler), event console (zentrap, zensyslog, zeneventd). On this scenario all devices send SNMP Traps and Syslog to both servers which are located in different geographical locations.

The second solution could be an Active Active more sophisticated, as the HA guide that you have referred with DRDB or similar and it permits provisioning configuration on two servers or more like I describe before.

Subject:	Zenoss 5.0.3 - SERVICED_MASTER_IP variable
Author:	[Not Specified]
Posted:	2015-10-23 07:26

Zenoss 5.0.3 - SERVICED_MASTER_IP variable