Hello everyone,
I am trying to setup monitoring over a high number of machines but I am not sure what is the best way to proceed. The situation is as follows:
a) We have around 100 different physical locations and there are around 30 devices that need to be monitored over each location
b) for each device there are at least 20 items that we need to be monitored
c) there are at least 5 items on each device that will be monitored through powershell scripts
d) We need to have fail over mechanism to be implemented
I have done some research but I could not find a comprehensive guideline on how to go implement so I thought I would write the different information that I have read and would appreciate any potential input I would get to correct the info:
1- We will need to deploy a proxy for each location, the proxy hardware requirements will not be very high as the load is low per each location
2- Install the zabbix servers and the databases on separate machines, recommended 3 servers and 3 databases to implement a failover cluster for each
3- 64 GB RAM will be required for the zabbix server
4- The bottleneck will be in the database.
Would like to know if the above approach (even though described briefly) would be suitable for our setup, and if the above makes sense what would the recommendations be related to the database being the bottleneck?
Regarding failover clustering, would it be recommended to use the standard zabbix HA configuration? or would it be advised to use something like pacemaker while maintaining same IP?
I am trying to setup monitoring over a high number of machines but I am not sure what is the best way to proceed. The situation is as follows:
a) We have around 100 different physical locations and there are around 30 devices that need to be monitored over each location
b) for each device there are at least 20 items that we need to be monitored
c) there are at least 5 items on each device that will be monitored through powershell scripts
d) We need to have fail over mechanism to be implemented
I have done some research but I could not find a comprehensive guideline on how to go implement so I thought I would write the different information that I have read and would appreciate any potential input I would get to correct the info:
1- We will need to deploy a proxy for each location, the proxy hardware requirements will not be very high as the load is low per each location
2- Install the zabbix servers and the databases on separate machines, recommended 3 servers and 3 databases to implement a failover cluster for each
3- 64 GB RAM will be required for the zabbix server
4- The bottleneck will be in the database.
Would like to know if the above approach (even though described briefly) would be suitable for our setup, and if the above makes sense what would the recommendations be related to the database being the bottleneck?
Regarding failover clustering, would it be recommended to use the standard zabbix HA configuration? or would it be advised to use something like pacemaker while maintaining same IP?
Comment