Upgrading your VPS plan — how it works, what you keep, what reboots

Resource changes on a LYLIX VPS happen entirely from the customer portal — no ticket, no maintenance window, no migration to a new server. This article walks the process end to end: which knob does what, what's preserved, what changes immediately, and what waits for your next reboot.

Two paths: configurable options vs. product change

LYLIX uses configurable options for resource scaling, not separate "small / medium / large" product tiers. There's one Linux VPS product and one PBX VPS product; you adjust the resources up or down within that product.

So when you "upgrade your plan" you're really:

1. Going to your service's product details page.
2. Clicking Upgrade / Downgrade Options.
3. Selecting new values for CPU, RAM, and/or Disk.
4. Confirming and paying the prorated difference.

You never switch to a "different product" — same product, new resource numbers.

Step by step

1. Sign in and find your service

Customer portal → click the service in your dashboard or under Services.

2. Click "Upgrade / Downgrade Options"

The button is in the action area of the product details page. You'll land on the configure-options form.

3. Review the warning (VM-based products)

On VM-based products (most LYLIX VPSes), the page shows an amber warning at the top:

Resource changes — please read before submitting:

• Disk space is one-way. You can increase disk size, but it cannot be reduced later. Disk sizes smaller than your current allocation are disabled below.
• CPU and RAM changes can move up or down, but the guest kernel only sees the new values after a reboot. You'll see a banner on the service page reminding you to reboot when convenient.

Read it. It's there because the most common support ticket about upgrades is "I downgraded the disk and now things broke." We prevent that at the cart UI; the warning explains why disk options below your current allocation are greyed out.

4. Pick your new resource levels

Each configurable option (CPU, RAM, Disk) has a dropdown of levels. The cart shows you each level's price increment relative to your current pricing.

• Disk dropdown: smaller-than-current options appear with "— downgrade not available" appended and are disabled. Pick equal-or-larger.
• CPU and RAM dropdowns: all options are selectable. Up or down.

5. Confirm the price summary

The cart shows you the prorated charge (or credit) and the new recurring price. Confirm.

6. Pay if there's a charge

For upgrades, the prorated invoice generates immediately and is charged to your card on file (or marked unpaid if you don't have one). For downgrades, no payment — you'll see the credit on the next invoice.

7. The hypervisor applies the change

This happens automatically within seconds:

• CPU and RAM: configured live on the VM. The hypervisor knows about the new allocation.
• Disk: grown live (you can verify with lsblk in the guest — the block device shows the new size).
• If CPU or RAM changed: a "pending reboot" custom field is set on your service. The portal product page shows an amber banner about it.

What survives the change

• Your IPv4 address
• Your IPv6 address (if you have one)
• Your hostname
• All snapshots
• All files on disk
• Your root password / SSH keys
• Reverse DNS configuration
• The service itself stays running through the upgrade (CPU/RAM hot-config doesn't bounce the VM)

Effectively: nothing on your VPS changes except how much CPU/RAM/disk it has available.

What requires a reboot before you see the change

From the guest OS perspective:

• CPU count: Linux kernels don't pick up new vCPUs at runtime reliably. nproc still shows the old count until reboot.
• RAM amount: same story. free -h still shows the old total until reboot.
• Disk size: the block device shows the new size right away (lsblk /dev/vda), but your partition and filesystem inside are still the old size. You need to grow the partition (growpart) and the filesystem (resize2fs / xfs_growfs) either online (if the tools support it) or after a reboot.

When to reboot

The portal banner doesn't force you. Reboot at whatever moment is least disruptive to whatever the VPS does. For a personal VPS, that's usually "tonight when nobody's using it." For a production service, "during your next maintenance window."

The reboot itself takes about 30 seconds. When the VM comes back, the banner clears and the guest OS sees the new resources.

How to reboot

1. Customer portal → your service.
2. Top action buttons: Reboot VM.
3. Confirm.

Verifying after the reboot

# SSH in (or use the browser console)
nproc          # new vCPU count
free -h        # new memory total
lsblk /dev/vda # new disk size

# If disk was grown, grow the filesystem too:
sudo growpart /dev/vda 1
sudo resize2fs /dev/vda1   # for ext4
# (or sudo xfs_growfs / for xfs)

If you change your mind

Resource upgrades are reversible (except disk).

• You upgraded CPU/RAM and want to scale back down: same flow — Upgrade/Downgrade Options, pick lower values, confirm. Downgrade credits to your account.
• You upgraded Disk and want to revert: you can't. Disk is one-way on VM-based VPSes. Either keep the larger allocation, or order a new smaller VPS and migrate.
• You bought an entirely wrong plan: contact billing within the first few days; we can often work something out. After significant usage, refund policy applies as written in the Terms of Service.

Container-based VPSes behave differently

If your VPS is on the legacy container-based platform (less common; older products), all three resources can be changed up and down live. No reboot, no disk-shrink restriction. The portal flow is the same; the underlying constraint is just absent.

You can tell which platform you're on by the management options shown on your service page — container-based VPSes show a different set of controls than VM-based VPSes. Most LYLIX VPSes ordered after ~2023 are VM-based.

Common questions

How long does the upgrade take to apply?
Hypervisor side: seconds. Guest side: until your next reboot for CPU/RAM; same for disk plus a filesystem resize.

Will my VM stop during the upgrade?
No. CPU/RAM hot-config keeps the VM running. Disk grow keeps it running.

Do I lose any data?
No. Resource changes don't touch the filesystem contents.

What if the prorated amount is wrong?
Open a billing ticket — we'll review. The proration is generated automatically; mistakes are rare but happen.

Can I upgrade just before renewal to lock in?
You can, and the proration handles the math correctly. There's no "discount for upgrading early" — pricing is the same whether you change on day 1 or day 29 of the cycle.