CRM crashed after building with go1.17

[gainsley]

After platform was built with go1.17 and deployed, we saw a crash in the CRM. Seeing this issue in 2022-02-03-1 load.

crmserver logs from automationBerlinCloudlet

fatal error: unreachable method called. linker bug

goroutine 180 [running]:
runtime.throw({0x2afbd97, 0x7f4c945dbf14})
  /usr/local/src/runtime/panic.go:1198 +0x71 fp=0xc0009ac648 sp=0xc0009ac618 pc=0xe88d71
runtime.unreachableMethod()
  /usr/local/src/runtime/iface.go:561 +0x25 fp=0xc0009ac668 sp=0xc0009ac648 pc=0xe5e6c5
golang.org/x/crypto/ssh.(*connectionState).readPacket(0xc0003f9d40, 0x7f4c945dbde0)
  /go/pkg/mod/golang.org/x/crypto@v0.0.0-20200622213623-75b288015ac9/ssh/transport.go:130 +0x42 fp=0xc0009ac6e8 sp=0xc0009ac668 pc=0x7f4c94b1dc82
golang.org/x/crypto/ssh.(*transport).readPacket(0xc0003f9d40)
  /go/pkg/mod/golang.org/x/crypto@v0.0.0-20200622213623-75b288015ac9/ssh/transport.go:114 +0x2c fp=0xc0009ac708 sp=0xc0009ac6e8 pc=0x7f4c94b1dbec
golang.org/x/crypto/ssh.(*handshakeTransport).readOnePacket(0xc0000c6f20, 0x1)
  /go/pkg/mod/golang.org/x/crypto@v0.0.0-20200622213623-75b288015ac9/ssh/handshake.go:374 +0x3d fp=0xc0009ac788 sp=0xc0009ac708 pc=0x7f4c94b0485d
golang.org/x/crypto/ssh.(*handshakeTransport).readLoop(0xc0000c6f20)
  /go/pkg/mod/golang.org/x/crypto@v0.0.0-20200622213623-75b288015ac9/ssh/handshake.go:197 +0x45 fp=0xc0009ac7c8 sp=0xc0009ac788 pc=0x7f4c94b03d25
golang.org/x/crypto/ssh.newClientTransport·dwrap·7()
  /go/pkg/mod/golang.org/x/crypto@v0.0.0-20200622213623-75b288015ac9/ssh/handshake.go:134 +0x26 fp=0xc0009ac7e0 sp=0xc0009ac7c8 pc=0x7f4c94b03b26
runtime.goexit()
  /usr/local/src/runtime/asm_amd64.s:1581 +0x1 fp=0xc0009ac7e8 sp=0xc0009ac7e0 pc=0x7f4c94638f81
created by golang.org/x/crypto/ssh.newClientTransport
  /go/pkg/mod/golang.org/x/crypto@v0.0.0-20200622213623-75b288015ac9/ssh/handshake.go:134 +0x21f

goroutine 1 [chan receive, 1 minutes]:
main.main()
  /go/src/github.com/mobiledgex/edge-cloud/cloud-resource-manager/cmd/crmserver/main.go:373 +0x14e5

goroutine 7 [chan receive]: k8s.io/klog.(*loggingT).flushDaemon(0xc00008e180)
  /go/pkg/mod/k8s.io/klog@v1.0.0/klog.go:1010 +0x6a
created by k8s.io/klog.init.0
  /go/pkg/mod/k8s.io/klog@v1.0.0/klog.go:411 +0xf4

angshuman@abhattacharjee-mac cloudlet % qmcc cloudlet show region=EU cloudlet=automationBerlinCloudlet

key:
organization: TDG
name: automationBerlinCloudlet
location:
latitude: 52.520007
longitude: 13.404954
ipsupport: Dynamic
numdynamicips: 254
state: CrmInitok <========
platformtype: Openstack
notifysrvaddr: 127.0.0.1:0
flavor:
name: DefaultPlatformFlavor
physicalname: berlin
containerversion: 2022-02-03-1
deployment: docker
crmaccesspublickey:
createdat: "2022-02-03T15:57:18.915690447+05:30"
defaultresourcealertthreshold: 80
gpuconfig:
driver:
name: nvidia-450
organization: TDG
secondarynotifysrvaddr: 127.0.0.1:0
dnslabel: automationberlincloudlet-tdg
rootlbfqdn: [shared.automationberlincloudlet-tdg.eu.mobiledgex.net](http://shared.automationberlincloudlet-tdg.eu.mobiledgex.net/)
angshuman@abhattacharjee-mac cloudlet %

80.187.131.187 is the IP of platformvm

The controller had also crashed once

angshuman@abhattacharjee-mac cloudlet % kubectl get pods -l app=controller
NAME READY STATUS RESTARTS AGE
controller-99ccc4cdc-dq4tf 1/1 Running 1 3h19m
controller-99ccc4cdc-tbtkt 1/1 Running 1 3h19m
angshuman@abhattacharjee-mac cloudlet %