Permormance optimiztion: avoid some list call, increase client-go QPS

[ZihanJiang96]

Issue

When we terminate a large mount of nodes at the same time, let's 600 nodes, lifecycle-manager can only process 75 node events per minute, which means 600/75=8 min. If we set the ASG Lifecycle hook's heartbeat timeout seconds to 300s, then some of the node events will never get processed and after the 300s timeout, the node will get terminated by ASG directly without proper drain, which leads to pod ungraceful shutdown.

Fixes/Improvements

1. Increase client-go QPS from 5 to 600, Burst from 10 to 2000
2. Add nodeMetadataMap to cache the node instanceID --> node name info (reduce the k8s client list call).
3. getNodeByName() is used during node drain/delete. it should use the Get() API instead of List()

After the improvements, lifecycle-manager could handle 200 node messages per minute.

[codecov[bot]]

Codecov Report

Attention: 12 lines in your changes are missing coverage. Please review.

Comparison is base (a70d012) 69.78% compared to head (308e7b1) 69.87%.

:exclamation: Current head 308e7b1 differs from pull request most recent head 3d43b6a. Consider uploading reports for the commit 3d43b6a to get more accurate results

Files	Patch %	Lines
pkg/service/server.go	57.14%	9 Missing and 3 partials :warning:

Additional details and impacted files

```diff @@ Coverage Diff @@ ## master #183 +/- ## ========================================== + Coverage 69.78% 69.87% +0.08% ========================================== Files 12 12 Lines 1314 1321 +7 ========================================== + Hits 917 923 +6 - Misses 325 326 +1 Partials 72 72 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.