Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.Original topic: tidb在哪个节点执行备份的命令为好？
|
username: TiDBer_5GvAkLi0
For the TiDB cluster, which node is best to execute the backup command?
|
username: TiDBer_5GvAkLi0
|
username: If the control machine is independent, then put it on the control machine, it won’t affect the TiDB nodes.
If the control node is not a separate machine, then choose the TiDB node with the most idle resources.
|
username: OK, thanks for the explanation. What about Dumpling? Is it also recommended to be placed on PD? It should be fine to place both BR and Dumpling on the control machine, right!? Because in my production environment, the TiDB-server and PD components are deployed on a high-configuration cloud server (without high availability), and I’m worried that placing the backup tools on the PD node might affect the server’s performance.
|
username: How large is the workload? If it’s not large, you can choose to handle it during off-peak hours. If the workload is large, you should handle it on PD during off-peak hours. If possible, it is recommended to ensure high availability.
Backup operations should still be carried out during off-peak periods.
Because it has just gone online and the business volume is not large, it will gradually increase later. Currently, there are 7 AWS cloud servers (TIDB-SERVER + PD 32C 64G 1 unit, TIKV 8C 61G 1900G 3 units, TIFLASH 16C 122G 3800G 2 units, MONITOR + control machine 8C 16G 1 unit). We will consider scaling up when the business volume increases in the future.
Is the QPS and TPS of the data high? If not, you can choose to operate on the PD node during off-peak hours.
At present, it doesn’t seem high. May I ask what the difference is between operating on PD and operating on the control machine? I currently want to back up on the control machine and make an NFS share on the control machine.
|
username: This topic was automatically closed 1 minute after the last reply. No new replies are allowed.