Testing ZFS with XCP-ng

borzel

some info about the o_direct flag: https://xenserver.org/blog/entry/read-caching.html

borzel

After reading some, I think we can not get it fully working in short time. Some of the tools might not work without o_direct support:

If the underlying file-system does not support O_DIRECT, utilities (e.g, vhd-util) may fail with error code 22 (EINVAL). Similarly, Xen may fail with a message as follows:
TapdiskException: ('create', '-avhd:/home/cklein/vms/vm-rubis-0/root.vhd') failed (5632 )

https://github.com/xapi-project/blktap/blob/master/README

Today there is just one fully working local ZFS configuration:

create your ZFS-Pool: zpool create ...
create a ZVOL (aka blockdevice): zfs create -V 50G pool/my_local_sr
create an EXT3 based SR: xe sr-create host-uuid=<UUID_of_your_host> type=ext shared=false name-label=<Name_of_my_SR> device-config:device=/dev/zvol/<pool-name>/<zvol-name>

It's not optimal, but working.

nraynaud

I am working on the issue right now. I am trying to exactly nail the problem.
there are 2 cases:

xe vdi-copy from another SR to a ZFS SR doesn't work
xe vdi-copy on the same SR doesn't work.

I that a complete assessment of the issues you found?

A cursory test seems to show that ssh ${XCP_HOST_UNDER_TEST} sed -i.bak 's/# unbuffered = true/unbuffered = false/' /etc/sparse_dd.conf solves the issue of intra-ZFS copies, but I am still confirming that I have not done anything else on my test box.

Thanks,
Nicolas.

borzel

@nraynaud said in Testing ZFS with XCP-ng:

xe vdi-copy from another SR to a ZFS SR doesn't work
xe vdi-copy on the same SR doesn't work.

Yes

Additional to that the copy from ZFS SR to another SR is not working.

nraynaud

ok, I really think changing /etc/sparse_dd.conf is the right path.

#!/usr/bin/env bash

# HOW TO create the passthrough: xe sr-create name-label="sda passthrough" name-description="Block devices" type=udev content-type=disk device-config:location=/dev/sda host-uuid=77b3f6ad-020b-4e48-b090-74b2a26c4f69

set -ex

MASTER_HOST=root@192.168.100.1
PASSTHROUGH_VDI=a74d267e-bb14-4732-bd80-b9c445199e8a

SNAPSHOT_UUID=19d3758e-eb21-f237-b8f7-6e2f638cc8e0
VM_HOST_UNDER_TEST_UUID=13ec74c2-9b57-a327-962f-1ebd9702eec4
XCP_HOST_UNDER_TEST_UUID=05c61e28-11cf-4131-b645-a0be7637c044
XCP_HOST_UNDER_TEST_IP=192.168.100.151
XCP_HOST_UNDER_TEST=root@${XCP_HOST_UNDER_TEST_IP}

INCEPTION_VM_UUID=a7e37541-fb9a-4392-6b54-60cf7ce3d08a
INCEPTION_VM_IP=192.168.100.32
INCEPTION_VM=root@${INCEPTION_VM_IP}

ssh ${MASTER_HOST} xe snapshot-revert snapshot-uuid=${SNAPSHOT_UUID}
NEW_VBD=`ssh ${MASTER_HOST} xe vbd-create device=1 type=Disk mode=RW vm-uuid=${VM_HOST_UNDER_TEST_UUID} vdi-uuid=${PASSTHROUGH_VDI}`
ssh ${MASTER_HOST} xe vm-start vm=${VM_HOST_UNDER_TEST_UUID}
until ping -c1 ${XCP_HOST_UNDER_TEST_IP} &>/dev/null; do :; done
sleep 20

# try EXT3
ssh ${XCP_HOST_UNDER_TEST} 'mkfs.ext3 /dev/sdb2 && echo /dev/sdb2 /mnt/ext3 ext3 >>/etc/fstab && mkdir -p /mnt/ext3 && mount /dev/sdb2 && df'
SR_EXT3_UUID=`ssh ${XCP_HOST_UNDER_TEST} "xe sr-create host-uuid=${XCP_HOST_UNDER_TEST_UUID} name-label=test-ext3-sr type=file other-config:o_direct=false device-config:location=/mnt/ext3/test-ext3-sr"`
TEST_EXT3_VDI=`ssh ${XCP_HOST_UNDER_TEST} xe vdi-create sr-uuid=${SR_EXT3_UUID} name-label=test-ext3-vdi virtual-size=214748364800`
TEST_VBD=`ssh ${XCP_HOST_UNDER_TEST} xe vbd-create device=1 type=Disk mode=RW vm-uuid=${INCEPTION_VM_UUID} vdi-uuid=${TEST_EXT3_VDI}`


ssh ${XCP_HOST_UNDER_TEST} reboot || true
sleep 20
until ping -c1 ${XCP_HOST_UNDER_TEST_IP} &>/dev/null; do :; done
sleep 20

ssh ${XCP_HOST_UNDER_TEST} xe vm-start vm=${INCEPTION_VM_UUID} on=${XCP_HOST_UNDER_TEST_UUID}
sleep 2
until ping -c1 ${INCEPTION_VM_IP} &>/dev/null; do :; done
sleep 20
ssh ${INCEPTION_VM} echo FROM BENCH
ssh ${INCEPTION_VM} 'apk add gcc zlib-dev libaio libaio-dev make linux-headers git binutils musl-dev; git clone https://github.com/axboe/fio fio; cd fio; ./configure && make&& make install'
ssh ${INCEPTION_VM} 'mkfs.ext3 /dev/xvdb && mount /dev/xvdb /mnt;df'
ssh ${INCEPTION_VM} 'cd /mnt;sync;/usr/local/bin/fio --name=randwrite --ioengine=libaio --iodepth=1 --rw=write --bs=4k --direct=1 --size=512M --numjobs=2 --runtime=30 --group_reporting' > ext3_write_result
ssh ${INCEPTION_VM} 'cd /mnt;sync;/usr/local/bin/fio --name=randwrite --ioengine=libaio --iodepth=1 --rw=write --bs=4k --direct=1 --size=512M --numjobs=2 --runtime=30 --group_reporting' >> ext3_write_result
ssh ${INCEPTION_VM} 'cd /mnt;sync;/usr/local/bin/fio --name=randwrite --ioengine=libaio --iodepth=1 --rw=read --bs=4k --direct=1 --size=512M --numjobs=2 --runtime=30 --group_reporting' > ext3_read_result
ssh ${INCEPTION_VM} 'cd /mnt;sync;/usr/local/bin/fio --name=randwrite --ioengine=libaio --iodepth=1 --rw=read --bs=4k --direct=1 --size=512M --numjobs=2 --runtime=30 --group_reporting' >> ext3_read_result
ssh ${XCP_HOST_UNDER_TEST} xe vm-shutdown uuid=${INCEPTION_VM_UUID}

# try ZFS
# install binaries that don't use O_DIRECT
rsync -r zfs ${XCP_HOST_UNDER_TEST}:
scp /Users/nraynaud/dev/xenserver-build-env/blktap-3.5.0-1.12test.x86_64.rpm ${XCP_HOST_UNDER_TEST}:
ssh ${XCP_HOST_UNDER_TEST} yum remove -y blktap
ssh ${XCP_HOST_UNDER_TEST} yum install -y blktap-3.5.0-1.12test.x86_64.rpm
ssh ${XCP_HOST_UNDER_TEST} yum install -y zfs/*.rpm
ssh ${XCP_HOST_UNDER_TEST} depmod -a
ssh ${XCP_HOST_UNDER_TEST} modprobe zfs
ssh ${XCP_HOST_UNDER_TEST} zpool create -f -m /mnt/zfs tank /dev/sdb1
ssh ${XCP_HOST_UNDER_TEST} zfs set sync=disabled tank
ssh ${XCP_HOST_UNDER_TEST} zfs set compression=lz4 tank
ssh ${XCP_HOST_UNDER_TEST} zfs list


SR_ZFS_UUID=`ssh ${XCP_HOST_UNDER_TEST} "xe sr-create host-uuid=${XCP_HOST_UNDER_TEST_UUID} name-label=test-zfs-sr type=file other-config:o_direct=false device-config:location=/mnt/zfs/test-zfs-sr"`
TEST_ZFS_VDI=`ssh ${XCP_HOST_UNDER_TEST} xe vdi-create sr-uuid=${SR_ZFS_UUID} name-label=test-zfs-vdi virtual-size=214748364800`
# this line avoids O_DIRECT in reads
ssh ${XCP_HOST_UNDER_TEST} "sed -i.bak 's/# unbuffered = true/unbuffered = false/' /etc/sparse_dd.conf"
# try various clone situations
ssh ${XCP_HOST_UNDER_TEST} xe vdi-copy uuid=${TEST_ZFS_VDI} sr-uuid=${SR_ZFS_UUID}
ssh ${XCP_HOST_UNDER_TEST} xe vdi-copy uuid=${TEST_ZFS_VDI} sr-uuid=${SR_EXT3_UUID}
ssh ${XCP_HOST_UNDER_TEST} xe vdi-copy uuid=${TEST_EXT3_VDI} sr-uuid=${SR_ZFS_UUID}

this script complete to the end without error.

nraynaud

If other people can reproduce my results, I propose to directly change the parameter in the XCP-ng distribution RPM.

olivierlambert

Your package is experimental, so feel free to add the modification inside it

borzel

@nraynaud said in Testing ZFS with XCP-ng:

If other people can reproduce my results

with the change in /etc/sparse_dd.conf I can copy my VMs from:

EXT3 -> ZFS
ZFS -> ZFS
ZFS -> EXT3

Yeha!

Thanks for your work!

By the way, my XCP-ng replication host at work is working just fine with ZFS-SR. All stable like ZFS should be.

olivierlambert

Yay!! Thanks for testing

eexodus

I have a clean install of 7.5 and I'm following the guide on the wiki but can't install zfs-test or enable the zfs module:

[root@xcp-ng-endlqfgb ~]# yum install --enablerepo="xcp-ng-extras" zfs-test
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
Resolving Dependencies
--> Running transaction check
---> Package zfs-test.x86_64 0:0.7.9-1.el7.centos will be installed
--> Processing Dependency: lsscsi for package: zfs-test-0.7.9-1.el7.centos.x86_64
--> Processing Dependency: ksh for package: zfs-test-0.7.9-1.el7.centos.x86_64
--> Processing Dependency: fio for package: zfs-test-0.7.9-1.el7.centos.x86_64
--> Processing Dependency: rng-tools for package: zfs-test-0.7.9-1.el7.centos.x86_64
--> Finished Dependency Resolution
Error: Package: zfs-test-0.7.9-1.el7.centos.x86_64 (xcp-ng-extras)
           Requires: fio
Error: Package: zfs-test-0.7.9-1.el7.centos.x86_64 (xcp-ng-extras)
           Requires: lsscsi
Error: Package: zfs-test-0.7.9-1.el7.centos.x86_64 (xcp-ng-extras)
           Requires: rng-tools
Error: Package: zfs-test-0.7.9-1.el7.centos.x86_64 (xcp-ng-extras)
           Requires: ksh
 You could try using --skip-broken to work around the problem
 You could try running: rpm -Va --nofiles --nodigest
[root@xcp-ng-endlqfgb ~]# zpool create tank /dev/sdb
The ZFS modules are not loaded.
Try running '/sbin/modprobe zfs' as root to load them.
[root@xcp-ng-endlqfgb ~]# /sbin/modprobe zfs
modprobe: FATAL: Module zfs not found.
[root@xcp-ng-endlqfgb ~]#

olivierlambert

@stormi this is for you

stormi

I knew we shouldn't have released on a Friday just before my holidays

XCP-ng community, note that I do that for you!

To answer the question, the zfs-tests package is not strictly needed so let's not install it (it pulls too many dependencies that we would have to add. We can consider making it installable later.). We forgot to add some steps in the procedure. We have updated it now.

If you run:
depmod -a

Then you should be able to run
modprobe zfs

This should bring you at least a step further.

borzel

@eexodus If it's not working, we can chat to troubleshoot your issue

stormi

@borzel It would be nice also to improve https://github.com/xcp-ng/xcp/wiki/ZFS-on-XCP-ng-7.5-and-later because I wrote it and I never used zfs!

eexodus

@stormi @borzel Thank you. I was able to create the zpool with those two extra commands. I then created a dataset and created a type=file and device-config:location SR. I realize ZFS is still a new feature, but is that the recommended way? I'm completely new to Xen so I am not familiar with creating SRs.

[root@xcp-ng-miqfcsgc] depmod -a
[root@xcp-ng-miqfcsgc] modprobe zfs
[root@xcp-ng-miqfcsgc] zpool create -f tank /dev/sda
[root@xcp-ng-miqfcsgc] zfs create tank/sr
[root@xcp-ng-miqfcsgc] xe sr-create host-uuid=MY_UUID name-label=zfs-sr type=file other-config:o_direct=false device-config:location=/tank/sr/zfs-sr

borzel

@eexodus said in Testing ZFS with XCP-ng:

zpool create -f tank /dev/sda

If you have 4k disks (aka Advanced Format) use ashift=12

zpool create -o ashift=12 -f tank /dev/sda

I would say, for the first testing, please use:

zfs set sync=disabled tank
zfs set compress=lz4 tank
zfs set atime=off tank

pool/discs can be live monitored with

zpool iostat -v 1

AllooTikeeChaat

@borzel Not sure if this thread is the right place to ask .. but whats the min and recommended requirements to test ZFS ?

olivierlambert

Well, depends on the level of perfs you want to achieve. A lot of RAM is the key.

borzel

@olivierlambert said in Testing ZFS with XCP-ng:

A lot of RAM is the key.

And if you want more performance, add more RAM

borzel

@stormi said in Testing ZFS with XCP-ng:

@borzel It would be nice also to improve https://github.com/xcp-ng/xcp/wiki/ZFS-on-XCP-ng-7.5-and-later because I wrote it and I never used zfs!

I have no write permissions