Installation on Synology Virtual Machine Managager

birca1987 · February 20, 2021, 11:20am

Is Open vSwitch enabled on your network card? My Nas is a 918+ and had to enable it in order to be able to connect to some add-ons on HassOS.

quart-temps · February 20, 2021, 11:28am

Sorry you mean on the NAS or on the VM ?

birca1987 · February 20, 2021, 11:36am

On the Nas in Network configuration it should have an option to enable Open vSwitch. When you install VMM in Synology it shows if Open vSwitch is enabled on Nas or not.

quart-temps · February 20, 2021, 12:19pm

Thank you. Yes it was enable.

spencer-scott · February 20, 2021, 1:36pm

Thanks for the suggestion. It was a simple SSL issue. I corrected it and I’m back up and running again! Thank you for all your hard work!

crosshelder · February 20, 2021, 2:12pm

Thank you so much for your work.
I have a DS418Play which is no suitable for VMM but i used the DS218+ VMM package and it is compatible. The only thing i had to do was to add this line ( support_ovs = “yes” ) to the file ( /etc/synoinfo.conf ) because of the activation of Open vSwitch.
Followed every step mentioned and imported your .ova file and voilá.

Restore complete snapshot and boommm. Perfect, with everything working.
Also have made an upgrade to synology ds418play memory to 8GB

Thanks for the work and cheers ( hope it´s cold enough )

chbndrhnns · February 21, 2021, 10:47am

This was the smoothest migration I have ever experienced. Now I am running VMM on my DS218+ with home assistant supervised, again.

DeltaNu1142 · February 21, 2021, 2:20pm

This guide is timely, because I was just looking at ways to migrate HA off of the RPi yesterday.

Using hassos_ova-5.11_syn-patch (1).ova, I get the VM running, increased storage to 64GB, get to the onboarding screen, and attempt to restore from this morning’s snapshot… but it fails each time. I see this screen when trying to connect to the VM from DSM:

Error

…and this when trying to connect via the browser (to 10.0.0.249:8123):

Error2

VM CPU usage is in the single digits. I can ping the new IP.

I suppose I could start configuring HA from scratch, but I’d rather not do that…

PatrickEhf · February 21, 2021, 5:31pm

What is the benefit of creating a new dedicated virtual switch instead of using the already existing “Default VM Network” virtual switch?

blavak68 · February 21, 2021, 10:09pm

did you enable USB during setup?

be more patient, this message is just about disabled USB

you need to change ip_syno to the ip your VM got or try http://homeassistant.local:8123

have you used e.g. duckdns? and do you have a snapshot made like this?

DeltaNu1142 · February 21, 2021, 10:23pm

I did in fact enable USB 3.0; and I was able to select my Z-wave dongle from the list in the drop-down below.

As for the IP, I was trying to connect to the VM address; my existing HA system is running on a RPi with IP 10.0.0.57.

EDIT: I use duckdns and my snapshot includes that. Everything seems to work fine until loading the snapshot. I have set up the VM at least four times now, each with the same result.

Can you explain this comment: “you need to change ip_syno to the ip your VM got…” I know the new IP assigned to the VM and am attempting to access it at port 8123 via the browser. I can ping that IP. Is there something else I need to do?

blavak68 · February 22, 2021, 11:19am

step 1 - you have created a new HA installation on VMM
the result?
connect to new HA via http://homeassistant.local:8123 ?

if the answer is yes, then:

step 2 - then try a snapshot from the old configuration

reassembling from snapshot may take longer, so be patient
if you have duckdns in the old snapshot, you also have it with the domain you are using on RPi, so try to open a port for ip_address_vmm:8123 on the router
for sure close the port for ip_rpi: 8123 or change it to others, eg: 8124, but before that you have to add to configuration.yaml (on RPi) next:

http:
server_port: 8124

then try logging in with your domain _ _ _. duckdns.org:8123 to new HA and old HA (RPi) should be go on address _ _ _. duckdns.org:8124

DeltaNu1142 · February 22, 2021, 11:56am

Well, that is the problem. I can connect to the new HA instance easily to restore my snapshot. After that, I can’t connect to the new HA instance at all. I get the error as shown in the browser window above.

I created a new HA installation on VMM last night. It’s assigned IP address 10.0.0.219. I connected to 10.0.0.219:8123 and selected my snapshot to restore. I left browser window open (unnecessary, I know) to restore overnight last night (figure 7 hours). In the morning, the browser still showed “restoring,” until just a few moments ago. I refreshed the browser–URL 10.0.0.219:8123–and it shows the same error as shown in my screen shot above.

I have shut down the VMM and restarted. It picks up the same IP address, but that doesn’t work either. Now, as HA is starting up on VMM, I get this page:

error

That’s pretty typical of HA while it’s loading up. I’ve seen that before when bringing up IP:8123 right after a restart. But before too long, I see the, “this page isn’t working” error.

Not sure where to go from here. The issue appears to be in my snapshot, but it’s working just fine on my RPi.

blavak68 · February 22, 2021, 12:13pm

and what the step 2 ?

settings ports on router …

my opinion is that after snapshot, HA on VMM is beating with the old duckdns setting and port settings on the router

Flop2006 · February 22, 2021, 12:20pm

Did you type the https instead of http?
That was a mistake I made.

DeltaNu1142 · February 22, 2021, 12:35pm

I’ll change my router port forwarding and see if that changes anything. What I don’t remember how to do is change the port on my RPi instance of HA… it’s not in config.yaml under http. I’m sure I did it through the UI somewhere, or it’s defaulted and the only place to change it is in config.yaml.

~~I figured since I’m only accessing locally, duckdns wouldn’t be a factor.~~
Ahhh… I’m up and running. There was some IP or port conflict, somehow, that was preventing my VM instance of HA from loading. Thank you.

Now I just need to find the path for the Z-wave USB on the Synology so I can change it in ZwaveJS.

EDIT: it’s /dev/ttyS1, but none of my devices are available, yet. Now I’m running into this error.

EDIT 2: OK, it’s working now! But with /dev/ttyUSB0? That’s not what dmesg grep | tty shows me. But… whatever. It’s working. My first Z-wave device took 3-4 seconds to respond, but everything else is much faster, so I’ll just give it some time, some updates, and some reboots to see if that fixes things. Thanks again! I bought @fredrike a beer and would like to do the same for you. Please let me know how.

blavak68 · February 22, 2021, 1:49pm

its OK, for me is OK when you run, maybe in future help you me

SergeS · February 22, 2021, 1:58pm

Hello guys,
I tried to move docker version to vmm version. I’ve got the first screen that HomeAssistant ready. Then I chose to restore from a snapshot. After that the only answer quite long ago: " homeassistant.local refused to connect.".
In Docker container I can open log or terminal and see what happens. Please advise how to track errors or the current conditions in the virtual machine? I pressed “Connect”, used “root” for the login and see “ha >”. What can I do now?

SergeS · February 22, 2021, 3:44pm

By the way, the processor load (DS 218+) is from 70 till 97% now constantly

DeltaNu1142 · February 22, 2021, 4:03pm

This sounds exactly like the issue I was having yesterday… see above.

@blavak68 correctly diagnosed my issue. Ultimately, the snapshot introduced the existing IP:port configuration from my old HA instance, which caused a conflict. The final fix for me was to remove the old HA instance from the network. It took some time and reboots for my router to purge the IP address and forwarding rules.