ipng/lab
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Add a virshall

Browse Source 
virshall can start, shutdown or destroy all VMs on a hypervisor LAB
so I can remove 'start' from create and pristine; sometimes it is useful
to only create (but not start) the LABs. Similarly, sometimes bringing
the LAB into pristine state, but without restarting the VMs, is useful.
This commit is contained in:
Pim van Pelt
2022-10-22 10:25:27 +02:00
parent 543074b72d
commit f1837236e2
4 changed files with 36 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
README.md
View File

@ -84,9 +84,12 @@ shut down the VMs and destroy the clones.
``` ```
lab:~/src/lab$ ./generate --host hvn0.lab.ipng.ch --overlay default lab:~/src/lab$ ./generate --host hvn0.lab.ipng.ch --overlay default
lab:~/src/lab$ lab=0 ./destroy ## remove VMs and ZFS clones lab:~/src/lab$ LAB=0 ./destroy ## remove VMs and ZFS clones
lab:~/src/lab$ lab=0 ./create ## create ZFS 'pristine' snapshot and start VMs lab:~/src/lab$ LAB=0 ./create ## create ZFS 'pristine' snapshot
lab:~/src/lab$ lab=0 ./pristine ## return the lab to the latest 'pristine' snapshot lab:~/src/lab$ LAB=0 ./virshall start ## Start the VMs
lab:~/src/lab$ LAB=0 ./virshall shutdown ## Gracefully stop the VMs
lab:~/src/lab$ LAB=0 ./virshall destroy ## Hard poweroff the VMs
lab:~/src/lab$ LAB=0 ./pristine ## return the lab to the latest 'pristine' snapshot
``` ```
### Generate ### Generate
@ -111,7 +114,7 @@ clones are destroyed. Obviously this is the most dangerous operation of the bunc
of the lab is that the VMs can be re-created off of a stable base image and a generated build. of the lab is that the VMs can be re-created off of a stable base image and a generated build.
``` ```
lab:~/src/lab$ lab=0 ./destroy ## remove VMs and ZFS clones on hvn0.lab.ipng.ch lab:~/src/lab$ LAB=0 ./destroy ## remove VMs and ZFS clones on hvn0.lab.ipng.ch
``` ```
### Create ### Create
@ -126,7 +129,7 @@ Typically, it's necessary to destroy/create, only when the build or the base ima
the lab can be brought back into a _factory default_ state by rolling back to the `pristine` snapshot. the lab can be brought back into a _factory default_ state by rolling back to the `pristine` snapshot.
``` ```
lab:~/src/lab$ lab=0 ./create ## create ZFS 'pristine' snapshots and start VMs lab:~/src/lab$ LAB=0 ./create ## create ZFS 'pristine' snapshots and start VMs
``` ```
### Pristine ### Pristine
@ -137,5 +140,6 @@ brought back into a default state by rolling back the disks to the `pristine` sn
the virtual machines. the virtual machines.
``` ```
lab:~/src/lab$ lab=0 ./pristine ## return the lab to the latest 'pristine' snapshot lab:~/src/lab$ LAB=0 ./pristine ## return the lab to the latest 'pristine' snapshot
lab:~/src/lab$ LAB=0 ./vishall start ## Start the VMs
``` ```

9
create
View File

@ -15,7 +15,7 @@
## * Unmount the staging directories ## * Unmount the staging directories
## * Start the VMs ## * Start the VMs
BASE=${BASE:=ssd-vol0/hvn0.chbtl0.ipng.ch/ssd-vol0/vpp-proto-disk0@20221019-release} BASE=${BASE:=ssd-vol0/hvn0.chbtl0.ipng.ch/ssd-vol0/vpp-proto-disk0@20221022-release}
BUILD=${BUILD:=default} BUILD=${BUILD:=default}
LAB=${LAB:=0} LAB=${LAB:=0}
@ -39,10 +39,3 @@ ssh root@$HVN "set -x; for node in 0 1 2 3; do VM=vpp${LAB}-\${node}; chown -R r
echo "* Unmounting and snapshotting pristine state" echo "* Unmounting and snapshotting pristine state"
ssh root@$HVN "set -x; for node in 0 1 2 3; do VM=vpp${LAB}-\${node}; umount $STAGING/\$VM; zfs snapshot ssd-vol0/\${VM}@pristine; done" ssh root@$HVN "set -x; for node in 0 1 2 3; do VM=vpp${LAB}-\${node}; umount $STAGING/\$VM; zfs snapshot ssd-vol0/\${VM}@pristine; done"
echo "* Starting VMs"
ssh root@$HVN "set -x; for node in 0 1 2 3; do VM=vpp${LAB}-\${node}; virsh start \$VM; done"
echo "* Committing OVS config"
scp overlays/$BUILD/ovs-config.sh root@$HVN:$LABDIR
ssh root@$HVN "set -x; LAB=$LAB $LABDIR/ovs-config.sh"

11
pristine
View File

@ -1,17 +1,10 @@
#!/bin/sh #!/bin/sh
BUILD=${BUILD:=default}
LAB=${LAB:=0} LAB=${LAB:=0}
## Do not touch below this line ## Do not touch below this line
LABDIR=/var/lab
STAGING=$LABDIR/staging
HVN="hvn${LAB}.lab.ipng.ch" HVN="hvn${LAB}.lab.ipng.ch"
## Bring back into pristine state ## Destroy the VMs and bring them back into pristine state
echo "* Restarting VMs from pristine snapshot" echo "* Restarting VMs from pristine snapshot"
ssh root@$HVN "set -x; for node in 0 1 2 3; do VM=vpp${LAB}-\${node}; virsh destroy \$VM; zfs rollback ssd-vol0/\${VM}@pristine; virsh start \$VM; done" ssh root@$HVN "set -x; for node in 0 1 2 3; do VM=vpp${LAB}-\${node}; virsh destroy \$VM; zfs rollback ssd-vol0/\${VM}@pristine; done"
echo "* Committing OVS config"
scp overlays/$BUILD/ovs-config.sh root@$HVN:$LABDIR
ssh root@$HVN "set -x; $LABDIR/ovs-config.sh"

23
virshall Executable file
View File

@ -0,0 +1,23 @@
#!/bin/sh
BUILD=${BUILD:=default}
LAB=${LAB:=0}
## Do not touch below this line
LABDIR=/var/lab
STAGING=$LABDIR/staging
HVN="hvn${LAB}.lab.ipng.ch"
CMD=$1
[ "$CMD" != "start" -a "$CMD" != "shutdown" -a "$CMD" != "destroy" ] && {
echo "Usage: LAB=0 ./virshall start|shutdown|destroy"
exit
}
echo "* Executing $CMD against all VMs on lab $LAB"
ssh root@$HVN "set -x; for node in 0 1 2 3; do VM=vpp${LAB}-\${node}; virsh $CMD \$VM; done"
[ "$CMD" = "start" ] && {
echo "* Committing OVS config"
scp overlays/$BUILD/ovs-config.sh root@$HVN:$LABDIR
ssh root@$HVN "set -x; $LABDIR/ovs-config.sh"
}