My Aurora RDS just peak for no reason

I have 2 of db.r5.12xlarge Aurora RDS (MySQL) [Engine Version: Engine version5.7.mysql_aurora.2.07.2] [Parameter Group: default.aurora-mysql5.7]

I define A (worked fine) and B (B have problem)

Previously used for a month without any problems.

But 1-2 days ago, I found too much response time from my Load Balance monitoring.

Then I checked monitor on db.r5.12xlarge Aurora RDS (B).

• 3 Hours https://i.imgur.com/dOMVuyA.png

• 2 Weeks https://i.imgur.com/V9faGpx.png

• 1 Week Performance Insight (Red Arrow start problem) https://i.imgur.com/RCahLzV.png

• 6 Hours Load Balance https://i.imgur.com/RIEcwLU.png

• 1 Hour Target Group (When getting normal) https://i.imgur.com/8rd0isP.png

• Benchmark connection time (on problem period) RDS A 0.0061 second, RDS B 0.3102 second.

I have checked on my EC2 every intances work normally and no CPU peak. I tried to add more instance to Load Balance. (But no help) I have checked Performance Insights no impact query. (Because I didn't change my code or anything.)

But my RDS A worked fine, My Application have to connect 2 Aurora RDS. (The work of both of them is almost the same.) And my RDS A work harder than (more queries, more insert, more select) RDS B (But no problem).

This happened to me once earlier this month. And all of a sudden it came back to normal without me doing anything about it. I really don't understand (This pic when RDS B back to normal https://i.imgur.com/kz5W4Wb.png).

Thank you very much and I'm really sorry about my bad english.

// ====

Update!

Problem day (No any special or routine operators in these days, everything same all days)

• 27 Nov - 3 Dem
• 17 Dem
• 27 Dem - 3 Jan
• 10 Jan