Project

General

Profile

Task #6872

Task #6869: Challenge OpenStack and OpenNebula with ucloud

Create ucloud-vm

Added by Nico Schottelius 4 months ago. Updated 3 months ago.

Status:
In Progress
Priority:
Normal
Assignee:
Target version:
-
Start date:
06/23/2019
Due date:
06/27/2019
% Done:

60%

Estimated time:
PM Check date:
06/26/2019

Description

This service will run on EVERY vm host and will do the heavy lifting / creating of VMs

For v1, implement the following:

  • Watch etcd on key prefix /v1/vm/
  • If there is a change, check whether
    • an existing VM on the current host should be killed
    • a new VM should be started
    • On event emit an email to sre@ (mail should be injected by env) -- just initially, later this will be disabled
  • Create a VM
    • if there is a new json entry in /v1/vm with a VM that is designated for this host and has not been started
    • Use qemu ... for it (sample below)
  • Delete a VM
    • If the status json entry of a VM is TO_STOP, stop it
    • After stopping was successful, put it into STOP
    • We need to register the runtime somewhere (v2)
  • Monitor a VM
    • If it is still running
    • If it is not running anymore, notify sysadmins
      • Write to /v1/log/
      • And/or send email to sre@
      • And restart the VM

For v2 implement:

  • Every 5 minutes poll etcd
  • Check whether the VMs that are supposed to run on this host are running
    • If one missing: send mail to sre@
  • Check that not OTHER VMs are running on this host
    • If there is one that doesn't belong here: send mail to sre@

Sample qemu commands from production

  • This is way too long, we can trim it down.
  • We will also use the -name attribute
  • We will also use ceph and smp
  • We will also use uuid
  • We will run VNC
    • v1: tcp
    • v4: unix sockets
oneadmin 3641289  0.5  0.1 2577776 407572 ?      Sl   Jun21  59:10 /usr/bin/qemu-system-x86_64 -name guest=one-24442,debug-threads=on -S -object secret,id=masterKey0,format=raw,file=/var/lib/libvirt/qemu/domain-24-one-24442/master-key.aes -machine pc-i440fx-2.8,accel=kvm,usb=off,dump-guest-core=off -m 1024 -realtime mlock=off -smp 1,sockets=1,cores=1,threads=1 -uuid d902efb7-2c66-4685-8914-95d38a717c90 -no-user-config -nodefaults -chardev socket,id=charmonitor,path=/var/lib/libvirt/qemu/domain-24-one-24442/monitor.sock,server,nowait -mon chardev=charmonitor,id=monitor,mode=control -rtc base=utc -no-shutdown -boot strict=on -device piix3-usb-uhci,id=usb,bus=pci.0,addr=0x1.0x2 -object secret,id=virtio-disk0-secret0,data=PR218Q61gRpXffWtRuIgQntokENHbAaBJGDDG66u3lg=,keyid=masterKey0,iv=XUHyWqKdHWGuva4eX+5qPw==,format=base64 -drive file=rbd:one/one-449-24442-0:id=libvirt:auth_supported=cephx\;none:mon_host=ceph1\:6789\;ceph2\:6789\;ceph3\:6789,file.password-secret=virtio-disk0-secret0,format=raw,if=none,id=drive-virtio-disk0,cache=none -device virtio-blk-pci,scsi=off,bus=pci.0,addr=0x4,drive=drive-virtio-disk0,id=virtio-disk0,bootindex=1 -drive file=/var/lib/one//datastores/100/24442/disk.1,format=raw,if=none,id=drive-ide0-0-0,readonly=on -device ide-cd,bus=ide.0,unit=0,drive=drive-ide0-0-0,id=ide0-0-0 -netdev tap,fd=35,id=hostnet0,vhost=on,vhostfd=37 -device virtio-net-pci,netdev=hostnet0,id=net0,mac=02:00:b3:39:79:75,bus=pci.0,addr=0x3 -vnc [::]:4162 -device cirrus-vga,id=video0,bus=pci.0,addr=0x2 -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x5 -msg timestamp=on

History

#1

Updated by William Colmenares 4 months ago

  • % Done changed from 0 to 60
  • Due date set to 06/27/2019

Added email on events, working in V2

#2

Updated by Jason Kim 4 months ago

  • PM Check date set to 06/26/2019
#3

Updated by William Colmenares 4 months ago

  • Status changed from New to In Progress

Working on v2, finding a way to display the vm ID on kill request

#4

Updated by Nico Schottelius 4 months ago

  • Assignee changed from William Colmenares to Ahmed Bilal
  • Status changed from In Progress to New
  • Description updated (diff)
#5

Updated by Ahmed Bilal 4 months ago

  • Instead of TO_STOP I think we have to react on REQUESTED_DELETE as mentioned here
  • Similarly, instead of TO_STOP --> STOP we need REQUESTED_DELETE --> DELETED
  • For VMs, which are killed in action (abnormally by some bug/error/user operation) we would set its status to "KILLED"
#6

Updated by Ahmed Bilal 4 months ago

  • Status changed from New to In Progress
#7

Updated by Nico Schottelius 4 months ago

I think we need to distinguish between DELETE and SUSPEND where

DELETE = VM is irrecoverable dead
SUSPEND = VM can be respawn and keeps IP and image

writes:

#8

Updated by Ahmed Bilal 4 months ago

  • Status changed from In Progress to Feedback
#9

Updated by Ahmed Bilal 3 months ago

  • Status changed from Feedback to In Progress

Also available in: Atom PDF