opennebula / addon-zfs Goto Github PK

Hi @kvaps in https://forum.opennebula.io/t/using-datastores-with-zfs/3075/7 you wrote:

Also you should know that the custom system storage feature has become available 
since version OpenNebula 5.0, and currently not developed for my zfs-addon.
If anyone want to develop it, need to write 2 new files:
/vmm/kvm/restore.zfs and /vmm/kvm/save.zfs

Can you tell me / us a slight bit more about that?

I would like to try to write them if it's useful & important.
But I don't exactly understand the role and how critical it is.

Which of these are true?

(*)As in the thread linked. I don't like it too much since it's a layering violation. ONE ends up the manager of the zvols.
i would rather not work on that even when granting that it might be someone who knows more than just the average zfs-forum-poster-gist can do with good procedures)

Hello
i can't create zfs datastore. Plz help

~ # lsb_release -a
LSB Version: core-9.20170808ubuntu1-noarch:security-9.20170808ubuntu1-noarch
Distributor ID: Ubuntu
Description: Ubuntu 18.04.4 LTS
Release: 18.04
Codename: bionic

~ # zfs list
NAME USED AVAIL REFER MOUNTPOINT
mzfs 2.45T 181G 25.4K /
mzfs/data 1.69T 181G 1.69T /mnt/data
mzfs/ftp 299G 181G 299G /mnt/ftp
mzfs/img 177G 181G 177G /mnt/img
mzfs/iso 16.7G 181G 16.7G /mnt/iso
mzfs/one 65.8K 181G 32.9K /mnt/one
mzfs/one/images 32.9K 181G 32.9K /mnt/one/images
mzfs/root 19.3G 181G 19.3G /
mzfs/swap 4.25G 182G 2.88G -

~ # cat /var/lib/one/remotes/datastore/zfs/zfs.conf
ZFS_CMD=/sbin/zfs
BRIDGE_LIST="localhost"
DATASET_NAME=mzfs/one/images
STAGING_DIR=/var/tmp

~ # cat ds.conf
NAME = "zfs"
DS_MAD = zfs
TM_MAD = zfs
DISK_TYPE = block
DATASET_NAME = mzfs/one/images

~ # onedatastore create ds.conf
ID: 113

log:
Fri May 8 01:42:56 2020 [Z0][ImM][I]: Command execution failed (exit code: 255): /var/lib/one/remotes/datastore/zfs/monitor PERTX0RSSVZFUl9BQ1RJT05fREFUQT48REFUQVNUT1JFPjxJRD4xMTA8L0lEPjxVSUQ+MDwvVUlEPjxHSUQ+MDwvR0lEPjxVTkFNRT5vbmVhZG1pbjwvVU5BTUU+PEdOQU1FPm9uZWFkbWluPC9HTkFNRT48TkFNRT56ZnM8L05BTUU+PFBFUk1JU1NJT05TPjxPV05FUl9VPjE8L09XTkVSX1U+PE9XTkVSX00+MTwvT1dORVJfTT48T1dORVJfQT4wPC9PV05FUl9BPjxHUk9VUF9VPjE8L0dST1VQX1U+PEdST1VQX00+MDwvR1JPVVBfTT48R1JPVVBfQT4wPC9HUk9VUF9BPjxPVEhFUl9VPjA8L09USEVSX1U+PE9USEVSX00+MDwvT1RIRVJfTT48T1RIRVJfQT4wPC9PVEhFUl9BPjwvUEVSTUlTU0lPTlM+PERTX01BRD48IVtDREFUQVt6ZnNdXT48L0RTX01BRD48VE1fTUFEPjwhW0NEQVRBW3pmc11dPjwvVE1fTUFEPjxCQVNFX1BBVEg+PCFbQ0RBVEFbL3Zhci9saWIvb25lLy9kYXRhc3RvcmVzLzExMF1dPjwvQkFTRV9QQVRIPjxUWVBFPjA8L1RZUEU+PERJU0tfVFlQRT4yPC9ESVNLX1RZUEU+PFNUQVRFPjA8L1NUQVRFPjxDTFVTVEVSUz48SUQ+MDwvSUQ+PC9DTFVTVEVSUz48VE9UQUxfTUI+MDwvVE9UQUxfTUI+PEZSRUVfTUI+MDwvRlJFRV9NQj48VVNFRF9NQj4wPC9VU0VEX01CPjxJTUFHRVM+PC9JTUFHRVM+PFRFTVBMQVRFPjxBTExPV19PUlBIQU5TPjwhW0NEQVRBW05PXV0+PC9BTExPV19PUlBIQU5TPjxDTE9ORV9UQVJHRVQ+PCFbQ0RBVEFbU0VMRl1dPjwvQ0xPTkVfVEFSR0VUPjxEQVRBU0VUX05BTUU+PCFbQ0RBVEFbbXpmcy9vbmUvaW1hZ2VzXV0+PC9EQVRBU0VUX05BTUU+PERJU0tfVFlQRT48IVtDREFUQVtCTE9DS11dPjwvRElTS19UWVBFPjxEU19NQUQ+PCFbQ0RBVEFbemZzXV0+PC9EU19NQUQ+PExOX1RBUkdFVD48IVtDREFUQVtOT05FXV0+PC9MTl9UQVJHRVQ+PFJFU1RSSUNURURfRElSUz48IVtDREFUQVsvXV0+PC9SRVNUUklDVEVEX0RJUlM+PFNBRkVfRElSUz48IVtDREFUQVsvdmFyL3RtcF1dPjwvU0FGRV9ESVJTPjxUTV9NQUQ+PCFbQ0RBVEFbemZzXV0+PC9UTV9NQUQ+PC9URU1QTEFURT48L0RBVEFTVE9SRT48L0RTX0RSSVZFUl9BQ1RJT05fREFUQT4= 110
Fri May 8 01:42:56 2020 [Z0][ImM][I]: /var/lib/one/remotes/datastore/zfs/../libfs.sh: line 503: RANDOM % 0: division by 0 (error token is "0")
Fri May 8 01:42:56 2020 [Z0][ImM][E]: Datastore template missing 'BRIDGE_LIST' attribute.
Fri May 8 01:42:56 2020 [Z0][ImM][E]: Error monitoring datastore 110: RGF0YXN0b3JlIHRlbXBsYXRlIG1pc3NpbmcgJ0JSSURHRV9MSVNUJyBhdHRyaWJ1dGUu. Decoded info: Datastore template missing 'BRIDGE_LIST' attribute.

I am trying to import Ubuntu app from marketplace
and get error
[ImageAllocate] Error allocating a new image. This Datastore only accepts persistent images.
what is the proper way to import images into ZFS datastore?

In relation with OpenNebula/one#6162 issue, we have updated the datastore drivers to take the driver message action through stdin. The change is rather simple, just read DRV_ACTION from stdin, see for example this change, something like:

DRV_ACTION=`cat -`
ID=$1

This needs to happen for every datastore/* action file and tm/monitor script (for example see here)

Note: Releases 6.4.x and 6.6.x will support both behaviors (configurable through oned.conf) to preserve compatibility

After following the instructions in the readme new VMs are not being created in the ZFS datastore. They are still created in the system default datastore.

Hi, can you please see if anything can be done, using ON 5.6 with local zfs (all in one box)
log:
Mon Nov 26 00:41:05 2018 [Z0][VM][I]: New state is ACTIVE
Mon Nov 26 00:41:05 2018 [Z0][VM][I]: New LCM state is PROLOG
Mon Nov 26 00:41:06 2018 [Z0][VM][I]: New LCM state is BOOT
Mon Nov 26 00:41:06 2018 [Z0][VMM][I]: Generating deployment file: /var/lib/one/vms/3/deployment.0
Mon Nov 26 00:41:07 2018 [Z0][VMM][I]: Successfully execute transfer manager driver operation: tm_context.
Mon Nov 26 00:41:08 2018 [Z0][VMM][I]: ExitCode: 0
Mon Nov 26 00:41:08 2018 [Z0][VMM][I]: Successfully execute network driver operation: pre.
Mon Nov 26 00:41:09 2018 [Z0][VMM][I]: Command execution fail: cat << EOT | /var/tmp/one/vmm/kvm/deploy '/var/lib/one//datastores/0/3/deployment.0' 'localhost' 3 localhost
Mon Nov 26 00:41:09 2018 [Z0][VMM][I]: error: Failed to create domain from /var/lib/one//datastores/0/3/deployment.0
Mon Nov 26 00:41:09 2018 [Z0][VMM][I]: error: internal error: process exited while connecting to monitor: 2018-11-26T05:41:09.027737Z qemu-system-x86_64: -drive file=/var/lib/one//datastores/0/3/disk.0,format=qcow2,if=none,id=drive-virtio-disk0,cache=none: Image is not in qcow2 format
Mon Nov 26 00:41:09 2018 [Z0][VMM][E]: Could not create domain from /var/lib/one//datastores/0/3/deployment.0
Mon Nov 26 00:41:09 2018 [Z0][VMM][I]: ExitCode: 255
Mon Nov 26 00:41:09 2018 [Z0][VMM][I]: Failed to execute virtualization driver operation: deploy.
Mon Nov 26 00:41:09 2018 [Z0][VMM][E]: Error deploying virtual machine: Could not create domain from /var/lib/one//datastores/0/3/deployment.0
Mon Nov 26 00:41:09 2018 [Z0][VM][I]: New LCM state is BOOT_FAILURE

oneadmin@localhost:~$ oneimage show 11
IMAGE 11 INFORMATION
ID : 11
NAME : ttylinux - kvm
USER : oneadmin
GROUP : oneadmin
LOCK : None
DATASTORE : zfs
TYPE : OS
REGISTER TIME : 11/26 00:39:46
PERSISTENT : No
SOURCE : localhost:tank/ONE/images/one-11
PATH : http://marketplace.opennebula.systems//appliance/4fc76a938fb81d3517000003/download/0
SIZE : 200M
STATE : used
RUNNING_VMS : 1

PERMISSIONS
OWNER : um-
GROUP : ---
OTHER : ---

IMAGE TEMPLATE
DEV_PREFIX="vd"
DRIVER="qcow2"
FORMAT="qcow2"
FROM_APP="27"
FROM_APP_MD5="b8ccdc63fb9d72ed12547fb1499c8b73"
FROM_APP_NAME="ttylinux - kvm"

VIRTUAL MACHINES

ID USER     GROUP    NAME            STAT UCPU    UMEM HOST             TIME
 3 oneadmin oneadmin ttylin-dev1     fail    0      0K localhost    0d 00h36

http://semver.org/

I added a zfs datastore according to the documentation (and also added a DS_MAD_CONF for zfs). After adding the datastore I'm getting a "Error monitoring datastore 100: Datastore driver 'zfs' not available" error.

Any idea where this error comes from? /var/tmp/one/datastore/zfs is available on the hypervisor.