Fully Open Edge Cloud

How to Install OS and Services on Rapid Space Server for a Partner

How a partner must install the OS and services on servers to become Rapid.Space hosting servers.
  • Last Update:2024-01-04
  • Version:003
  • Language:en

Here is the procedure to install a OCP server that will be used as Rapid.Space hosting server.

1. Install OS

Recommended OS is Debian 12. Debian 11 is also supported. You should follow rapidspace-HowTo.Partner.Server.OS.Installation to install the OS.

Install the OS on the small SSD disk so that the big SSD disk can be used totally by the Rapid.Space VM.

2. Install Production re6st

Rapid.Space Operator will give you one time token for re6stnet installation.

In order to request the token navigate to panel.rapid.space --> Tickets --> Add.

Use Ticket Type "Information", in "Subject" type "Production re6st new token" and in the "Your Message" write:

I need a new Production re6st token for machine named "name", which will be hosted in "DC name" datacenter.

Where the "name" and "DC name" are important information for the Rapid.Space Operator.

When Rapid.Space Operator will generate token for you, the given ticket will be closed, and you'll find the token in the reply.

With this token in hand you can start installing the re6st on the machine:

wget https://deploy.rapid.space/re6st
bash re6st

It will ask you for the token.

Then you need to configure re6st by editing /etc/re6stnet/re6stnet.conf and add this line

interface XXXXX

with XXXXX being the physical network interface name of your server.

Note: In case you want to have a free re6stnet on your own PC instead of the OCP servers, you can follow rapidspace-HowTo.Use.SDN.For.A.Single.Server.

 

 

3. Install SlapOS

Rapid.Space Operator will give you one time token for a computer installation.

In order to request the token navigate to panel.rapid.space --> Tickets --> Add.

Use Ticket Type "Information", in "Subject" type "Production computer new token" and in the "Your Message" write:

I need a new Production computer token for machine named "name", which will be hosted in "DC name" datacenter.

Where the "name" and "DC name" are important information for the Rapid.Space Operator.

When Rapid.Space Operator will generate token for you, the given ticket will be closed, and you'll find the token in the reply.

With this token in hand you can start installing the vifib on the machine:

wget https://deploy.rapid.space/vifib
bash vifib

Please note that this script will also install a cron called "upgrader" to upgrade the machine regularly. It will also install the SSH public keys of all the vifib team in root account so that the team can SSH into your server.

Now reboot the server:

reboot

After the reboot, ensure that KVM is allowed in nested mode with:

grep -q 'Y\|1' /sys/module/kvm_*/parameters/nested && echo "KVM nested OK" || echo "KVM nested ERROR"

Which shall return

KVM nested OK

4a. Set the disk to be used by the VM - VPS variant

Locate the big SSD disk with using lsblk:

lsblk -pd -e 11

The example output is:

NAME     MAJ:MIN RM   SIZE RO TYPE MOUNTPOINT
/dev/sda   8:0    0   3.5T  0 disk
/dev/sdb   8:16   0 223.6G  0 disk

From this output one can see that /dev/sda is the big disk.

Allow usage of that disk by appending to the /etc/opt/slapos/slapos.cfg:

[manager:devperm]
allowed-disk-for-vm = /dev/sda

And in the [slapos] section of /etc/opt/slapos/slapos.cfg add entry:

manager_list = devperm,prerm

Link the disk in the configuration:

ln -sf /dev/sda /dev/disk/disk-path-for-vm

This will result in using this disk for VM instantiated the server.

Make the link persist on reboot by:

echo "@reboot root ln -sf /dev/sda /dev/disk/disk-path-for-vm" > /etc/cron.d/link-disk-path-for-vm

4b. Set the disk to be used by the VM - VPSbrute variant

If the machine is Capri server which shall hold VPS brute configuration you can use deploy script like:

rm -f rapid.space-vps-brute-capri ; wget https://deploy.erp5.net/rapid.space-vps-brute-capri ; bash rapid.space-vps-brute-capri

5. Do the stress test

Before putting the machine in the cloud, the stress testing has to be done.

Install stress-ng:

apt -y install stress-ng

Run in for 24 hours:

nohup stress-ng --cpu 60 --timeout 24h </dev/null >/dev/null 2>&1 &

6. Add the machine to SlapOS Cloud by Rapid.Space administrator

Ask the rapid.space administrator to perform those steps.

Prerequsities:

Locate the server created in 3. Install SlapOS, remember it's reference (COMP-XXX).

Login as "rapid.space" user in https://vifib.erp5.net/ and locate the computer by its reference, then:

  • in Capacity tab set Capacity to 40
  • in Model tab set Network to proper network

Login as "rapid.space" user in https://slapos.vifib.com/ and locate the computer by its reference, then

  • set Monitoring: Enabled
  • set Upgrade: Auto Upgrade
  • set Allocation Scope: Open for Subscribers Only
  • by using Supply
    • install the <KRV>
  • by using Transfer:
    • set Future Location: <DHC>
    • Future Project: Rapid Space Computers