Closed hakman closed 4 years ago
I0505 15:16:54.605527 8984 volumes.go:86] AWS API Request: ec2metadata/GetToken
I0505 15:16:54.606897 8984 volumes.go:86] AWS API Request: ec2metadata/GetDynamicData
I0505 15:16:54.608283 8984 volumes.go:86] AWS API Request: ec2metadata/GetMetadata
I0505 15:16:54.609035 8984 volumes.go:86] AWS API Request: ec2metadata/GetMetadata
I0505 15:16:54.609834 8984 volumes.go:86] AWS API Request: ec2metadata/GetMetadata
I0505 15:16:54.614547 8984 main.go:279] Mounting available etcd volumes matching tags [k8s.io/etcd/events k8s.io/role/master=1 kubernetes.io/cluster/k8s.test.com=owned]; nameTag=k8s.io/etcd/events
I0505 15:16:54.616581 8984 volumes.go:86] AWS API Request: ec2/DescribeVolumes
I0505 15:16:54.829148 8984 mounter.go:294] Trying to mount master volume: "vol-0c276c2e24d65df14"
I0505 15:16:54.829310 8984 volumes.go:86] AWS API Request: ec2/AttachVolume
W0505 15:16:55.102292 8984 volumes.go:341] Invalid value '/dev/xvdu' for unixDevice. Attachment point /dev/xvdu is already in use
I0505 15:16:55.102394 8984 volumes.go:86] AWS API Request: ec2/AttachVolume
I0505 15:16:55.412977 8984 volumes.go:347] AttachVolume request returned {
AttachTime: 2020-05-05 15:16:55.384 +0000 UTC,
Device: "/dev/xvdv",
InstanceId: "i-08d1d1f8a9ebe3409",
State: "attaching",
VolumeId: "vol-0c276c2e24d65df14"
}
I0505 15:16:55.413125 8984 volumes.go:86] AWS API Request: ec2/DescribeVolumes
I0505 15:16:55.537078 8984 mounter.go:308] Currently attached volumes: [0xc0002b0000]
I0505 15:16:55.537099 8984 mounter.go:72] Master volume "vol-0c276c2e24d65df14" is attached at "/dev/xvdv"
CC @justinsb
Some nits for more verbose logging and comments, but this looks good - thank you!
/approve /lgtm
Trying to attach volumes with same device names adds unnecessary delays. With this fix, kube-apiserver starts in less than a minute after NodeUp finishes.
Fixes #217