1. fix typos;
2. fix clusterrole rules;
3. use os package instead of deprecated io/ioutil;
4. update go modules.

fix: check np switch

fix: do not reuse released ip after subnet updated

fix nat rule

When netpol is added to a workload, the workload's POD can be accessed using service

Co-authored-by: wang_yudong <wang_yudong@inspur.com>

when update subnet's execpt ip,we should filter repeat ip

update: update 1.7-1.8 script

perf: do not send traffic to ct if not designate to svc

fix: add back the leader check

Previous patch doesn't change the readiness prob interval which may leader two probs run in the same period and always steal the lock from ovn-northd. Increase the interval and add back the leader check.

update Go modules

sync live migration vm port

fix port_security

update delete operation for statefulset pod

chore: update klog to v2 which embed log rotation

fix: add kube-ovn-cni prob timeout

When node load is high, the nc probe may take more than the default 1 second and lead prob failed.

append add db compact for nb and sb db

deleting all chassises which are not nodes

add db compact for nb and sb db

add vendor param for fix list LR

fix LB: skip service without cluster IP