fix: Add target ping interval

[destinyoooo]

This PR includes the following improvements:

1. Add target ping interval ：

• Added pingInterval field in target.go, defaulting to 30 seconds
• Implemented ping caching mechanism, only ping the database when the last ping was more than pingInterval ago
• Reduced unnecessary ping operations to the database, especially when using cached collectors

[burningalchemist]

Hey @destinyoooo, I have a small question here:

30 seconds is a hardcoded parameter. Do you think there are scenarios where it needs to be adjusted by the user?

Currently, one of the options to reduce that constant pinging (which isn't good for data warehouse solutions cost-wise) is to use the flag to disable ping completely (at a cost of potential connection losses).

So I'm wondering whether we need to expose pingInterval in configuration, or it's a sane default to have. What do you think?

[destinyoooo]

Hey @destinyoooo, I have a small question here:

30 seconds is a hardcoded parameter. Do you think there are scenarios where it needs to be adjusted by the user?

Currently, one of the options to reduce that constant pinging (which isn't good for data warehouse solutions cost-wise) is to use the flag to disable ping completely (at a cost of potential connection losses).

So I'm wondering whether we need to expose pingInterval in configuration, or it's a sane default to have. What do you think?

Hello! I've completed the modification to change pingInterval from a hardcoded 30 seconds to a configurable parameter. Here are the related improvements:

1. Added configuration option : Added a ping_interval parameter in the global configuration section, allowing users to adjust it as needed.

2. Default value setting : Set the default value to 0, which means ping every time, maintaining the same behavior as before.

3. Flexible configuration :

   • Set to 0: Ping every time during scrape
   • Set to a value greater than 0: Ping at the specified interval, e.g., 30s means ping every 30 seconds

[destinyoooo]

In my opinion. This improvement is very necessary because:

• For services like data warehouses that charge per query, reducing unnecessary ping operations can lower costs
• Database performance and stability vary across different environments, and users may need to adjust the ping interval based on actual conditions
• It provides a middle ground between completely disabling ping and pinging every time

[burningalchemist]

Yeah, makes sense to me. 👍 Let me have a brief look and we merge it soon. 👍

[burningalchemist]

There is a minor nitpick on the comment, otherwise looks good to me 👍

[burningalchemist]

Please adjust the comment for the default value 🙂