Need to add retry, when synchronization

[tao12345666333]

I occasionally encounter failures when requesting the admin API of the APISIX cluster.

{"error_msg":"timeout"} 

This will cause the Ingress Controller to fail to create the corresponding resources normally. like this:

$ kubectl -n cloud logs ingress-controller-d978b79d4-s8kgk  
E1203 15:43:38.194554       1 upstream.go:26] list upstreams in etcd failed, group: apisix-gw-lb.infraop.svc:9080, err: json转换失败
E1203 15:43:38.196041       1 builder.go:154] solver upstream failed, find upstream from etcd failed, upstream: &{ID:<nil> FullName:0xc000725450 Group:0xc0007253c0 ResourceVersion:0xc0007253d0 Name:0xc000725420 Type:0xc000725460 HashOn:<nil> Key:<nil> Nodes:[0xc00092a0e0] FromKind:<nil>}, err: list upstreams failed, err: json转换失败
E1203 15:56:54.560663       1 upstream.go:26] list upstreams in etcd failed, group: apisix-gw-lb.infraop.svc:9080, err: json转换失败
E1203 15:56:54.560703       1 builder.go:154] solver upstream failed, find upstream from etcd failed, upstream: &{ID:<nil> FullName:0xc000724710 Group:0xc000724650 ResourceVersion:0xc000724670 Name:0xc0007246d0 Type:0xc000724720 HashOn:<nil> Key:<nil> Nodes:[0xc000b1a800] FromKind:<nil>}, err: list upstreams failed, err: json转换失败
E1203 16:08:28.243429       1 upstream.go:26] list upstreams in etcd failed, group: , err: json转换失败
E1203 16:08:28.243469       1 builder.go:154] solver upstream failed, find upstream from etcd failed, upstream: &{ID:<nil> FullName:0xc000b471c0 Group:0xc000b47160 ResourceVersion:0xc000b47170 Name:0xc000b47180 Type:0xc000b471d0 HashOn:<nil> Key:<nil> Nodes:[0xc000bb9f80] FromKind:0xc000b471a0}, err: list upstreams failed, err: json转换失败
E1203 16:10:57.544821       1 upstream.go:26] list upstreams in etcd failed, group: , err: json转换失败
E1203 16:10:57.544850       1 builder.go:154] solver upstream failed, find upstream from etcd failed, upstream: &{ID:<nil> FullName:0xc00028ab90 Group:0xc00028aa70 ResourceVersion:0xc00028aa80 Name:0xc00028aa90 Type:0xc00028aba0 HashOn:0xc00028abb0 Key:0xc00028abc0 Nodes:[0xc000b20820] FromKind:0xc00028ab30}, err: list upstreams failed, err: json转换失败
E1203 16:45:40.909181       1 upstream.go:26] list upstreams in etcd failed, group: apisix-gw-lb.infraop.svc:9080, err: json转换失败
E1203 16:45:40.909220       1 builder.go:154] solver upstream failed, find upstream from etcd failed, upstream: &{ID:<nil> FullName:0xc0000ff0a0 Group:0xc0000fed70 ResourceVersion:0xc0000feda0 Name:0xc0000fedf0 Type:0xc0000ff0b0 HashOn:<nil> Key:<nil> Nodes:[0xc000b7a840] FromKind:<nil>}, err: list upstreams failed, err: json转换失败
E1203 16:45:41.051755       1 upstream.go:26] list upstreams in etcd failed, group: , err: json转换失败
E1203 16:45:41.051784       1 builder.go:154] solver upstream failed, find upstream from etcd failed, upstream: &{ID:<nil> FullName:0xc00077e3d0 Group:0xc00077e370 ResourceVersion:0xc00077e380 Name:0xc00077e390 Type:0xc00077e3e0 HashOn:<nil> Key:<nil> Nodes:[0xc000b7ab20] FromKind:0xc00077e3b0}, err: list upstreams failed, err: json转换失败
E1203 16:45:41.137569       1 upstream.go:26] list upstreams in etcd failed, group: , err: json转换失败
E1203 16:45:41.137605       1 builder.go:154] solver upstream failed, find upstream from etcd failed, upstream: &{ID:<nil> FullName:0xc0004a4d40 Group:0xc0004a4cb0 ResourceVersion:0xc0004a4cc0 Name:0xc0004a4cd0 Type:0xc0004a4d50 HashOn:0xc0004a4d80 Key:0xc0004a4d90 Nodes:[0xc000533520] FromKind:0xc0004a4d30}, err: list upstreams failed, err: json转换失败

[gxthrj]

Yes ， retry is necessary feature.
We can support this feature first.

[tokers]

We may re-add this failure job to the queue.

[tokers]

Assigned to me.