Hello
curious it doesn’t work…
Vultr support said
"
Hi
Certain custom installations need uefi disabled in the hypervisor file, which is what I did to your server.
"
regards
Aha, another host for fusionhub that is potentially under $5/mo. I probably didn’t find this in the past because they don’t appear to have servers in Chicago, which is where I would prefer to host. I have several affordable VPSs with a few different companies in Chicago but none of them support the ISO that Pep provides. I wish there were a way to create a container myself (Docker for instance) to run this.
Anyway, do you need the 2GB DreamCompute or will the 512KB one work?
any updates on this issue?
Has anyone been able to figure this out? i also followed martins video which ended with me being stuck with not being able to access the webgui after getting an ip address from vultr.
I heard that we now need to submit a ticket to Vultr support to turn UEFI off on the vm, then the raw image will work. If you then save a snapshot future vultr vms deployed from the snapshot will also have UEFI disabled.
Certainly my original Fusionhub snapshot still works as expected.
hi martin,
so i contacted vultr to submit a ticket to have UEFI disabled they set my console to bios according to them. it seems i still cant access the the webgui.
Do this:
I’m having the same issues with Vultr.
I started with the 8.0.1 RAW image, had them disable UEFI.
Now I am stuck at “No bootable device.”
I tried to reinstall, same issue.
Do I need to start with the version 6 RAW image, and then upgrade to 8?
Select Cloud Compute - Shared CPU.
Pick a location.
Choose Image. Be sure to choose ‘Snapshot’.
Add New
Upload snapshot.
Instructions courtesy of aquablue
Following q8reflex’s post above, you go to the Snapshot tab at vultr, not the ISO tab. On the bottom of the snapshot tab you can enter the URL for the contents of the zip file linked above, but you have to unzip it and upload it somewhere first. You can use dropbox or if you have signed up for vultr and don’t have another online storage or hosting spot, you can create a new temporary instance.
https://download.peplink.com/firmware/fusionhub/fw-fusionhub-7.1.1-build1538-raw.zip
Seems like checking UEFI is a good idea.
Choose Plan. All you need is ‘Regular Cloud Compute.
If you want to save a buck, unselect ‘Auto Backups’
Thanks for the steps, but those are exactly what I am doing.
I still end up failing to boot.
I’ve tried with 6.x, 7.x and 8.x.
When you took these images, did it successfully create for you?
I’ve tried in NY/NJ and Atlanta with the same results.
Also I’ve tried with and without UEFI for the snapshot upload…makes no difference.
1 Like
Just FWIW, it’s not “you.” We had two running instances with Vultr at Atlanta. Both died with errors such as you see. We spent hours trying to get them going again. We eventually destroyed them and are now looking for another solution. One of our customers, very technically savvy, had the exact same experience.
2 Likes
From What I see it is a subtle change in how VULTR is uploading the raw images.
I had an 8.1.0 snapshot. and I was able to create a new machine from that (no playing with UEFI settings) but a newly uploaded 8.0.1 has the same error. I’m guessing that there is a differece in the upload code.
It is definately booting into UEFI.
If I have time I will try and mount each snapshot into a liveimage and look for the subtle changes.
3 Likes
Interesting and useful. In the cases we have seen the instances were working perfectly and had been for years (literally.) Then Vultr “changed something.”
2 Likes
After doing a number of tests, the issue is within VULTR… The images on the disk are correct, so it must have been some other infro in a regular snapshot. I had support disable UEFI and that wasn’t enough… they then disabled something else and it worked… I have requested exactly how to request this specific config, and will add this to this thread.
1 Like
I have The feedback
You must disable UEFI via support ticket. There seems to be two places they have to change the settings so if it doesn’t work the first time tell them to check again and reference DPW-79PBB that an additional change was needed.
3 Likes
Thank you, Paul. That was very helpful. Too bad they could/would not fix it for us after we submitted a ticket.
1 Like
I reached out about this too and made sure to reference ticket DPW-79PBB. They got back to me within minutes and now fusionHub is up and running. Thank you @Paul_Mossip for getting back with that ticket number, I’m sure this helped a ton!
If you make a snapshot of the 8.0.1 image before you configure the license, then you will be able to restore a clean original image with the archived boot settings for future installs.
1 Like
Thanks. Just referenced the ticket number and they set both the instance and the 8.0.1 snapshot in my account to legacy. I had to reinstall the instance from the snapshot and then everything booted fine,.
How are you all assigning a FusionHub license using this method?
I can get the instance to boot using a mix of the steps above for Vultr, DO and Linode, but once I get to the setup wizard I cannot add a license key as I get the “VMware model mismatch. Please check VMware Model and Server Address.” error.
The previous workaround was this was downgrade to 7.0.1 or lower but this firmware no longer seems to be available from the peplink download site.