Create disk snapshot for VolumeSnapshot object too slowly(> 5min)

[borgerli]

What happened:
We have lots of existing volume snapshots(> 15,000) in our cluster. Sometimes we created a VolumeSnapshot object , the real snapshotting operation took place after a long time, such more than 5 minutes.

What you expected to happen:
no delay or very short delay
How to reproduce it:
This issue needs lots of snapshot objects and not easy to reproduce in test clusters.
Anything else we need to know?:
cause by unnecessary k8s cache resync
Environment:

• Driver version:
• Kubernetes version (use kubectl version):
• OS (e.g. from /etc/os-release):
• Kernel (e.g. uname -a):
• Install tools:
• Others: