Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.
Original topic: Prometheus监控采集问题
[TiDB Usage Environment] Production Environment
[TiDB Version] 6.1.6
[Encountered Problem: Problem Phenomenon and Impact]
The TiDB cluster is on physical machines, using a separate switch for internal communication. A separate virtual machine was set up for monitoring, but it cannot pull monitoring data. In this situation, should I consider using the Push Gateway method to collect data, or is it better to deploy the monitoring on one of the physical machines in the cluster, such as on TiFlash?
Finally, based on the official recommendation, it was placed on the PD machine.
If there are extra machines, I still think it’s better to deploy TiUP and monitoring on separate machines. Monitoring can sometimes take up a lot of space.
TiUP doesn’t matter, you can put them together, but make sure to do backups properly. The main issue with monitoring is not the machine, but network disconnection. This seems unsolvable, so you’ll have to find other solutions.
Network issues cannot be resolved by deployment location. After all, this is a cluster database, and unless it is installed in standalone mode, it will be affected by network issues.
The TiDB cluster consists of 16 physical machines, each equipped with 4 network cards: 2 for business use and 2 for internal communication. They use a separate switch, so cluster communication will not be affected.
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.
|
username: 啦啦啦啦啦
|
username: