Dear Zabbix Community,
I am about to migrate a medium Zabbix environment (~ 1500 Hosts and as many NVPS tendency growing) from a VM Deployment to a containerized setup for various reasons.
Currently I am still in the PoC phase and so far setup Server + DB as well as WebFE with docker compose.
- Zabbix Server 5.2
- DB Postgres Timescale DB 2.0.1
- "Zabbix" WebFE 5.2
The general idea is to set up several proxies (according to a reasonable mapping to the organizational structure of the company I am doing it for) which should be responsible
for the data collection and buffering, whereas the server should be used for item computing and persisting to the DB.
I am about to go for the containerized proxy setup and have seen only sqlite and mysql is supported with the zabbix provided proxy docker images
As my database know-how is unfortunately only superficial, I was wondering if the mysql instances for the proxy DBs can actually keep up with the timescale
DB of the server when the system scales overt time. I spent a bit of time reading the promising performance stats of timescale with large environments but have no
time to setup a testsystem to simulate how large datasets would compare between mysql and timescale DBs and therefore have to trust what I was able to research.
So in this regard, has anyone of you made experiences with a similar setup and has any lessons learned that would be worth to share?
Also has anyone of you found a zabbix proxy container image which would support postgres?
I am looking forward for your suggestions.
I am about to migrate a medium Zabbix environment (~ 1500 Hosts and as many NVPS tendency growing) from a VM Deployment to a containerized setup for various reasons.
Currently I am still in the PoC phase and so far setup Server + DB as well as WebFE with docker compose.
- Zabbix Server 5.2
- DB Postgres Timescale DB 2.0.1
- "Zabbix" WebFE 5.2
The general idea is to set up several proxies (according to a reasonable mapping to the organizational structure of the company I am doing it for) which should be responsible
for the data collection and buffering, whereas the server should be used for item computing and persisting to the DB.
I am about to go for the containerized proxy setup and have seen only sqlite and mysql is supported with the zabbix provided proxy docker images
As my database know-how is unfortunately only superficial, I was wondering if the mysql instances for the proxy DBs can actually keep up with the timescale
DB of the server when the system scales overt time. I spent a bit of time reading the promising performance stats of timescale with large environments but have no
time to setup a testsystem to simulate how large datasets would compare between mysql and timescale DBs and therefore have to trust what I was able to research.
So in this regard, has anyone of you made experiences with a similar setup and has any lessons learned that would be worth to share?
Also has anyone of you found a zabbix proxy container image which would support postgres?
I am looking forward for your suggestions.
Comment