Part 1 - Overview & Concepts
What ServersCTL hosting pools are
ServersCTL (serversctl.com) is the control plane for redundant server infrastructure: enrol Linux VMs with the ServersCTL agent, monitor heartbeats, cut over DNS between peers, run stack backups, and (on the cPanel preset) orchestrate account replication and live WHM transfers.
A server pool is one deployment in your dashboard — a set of members sharing failover DNS and pool-level settings. Members run what you actually install on each host. The dashboard exposes member tabs for OpenLiteSpeed, MariaDB/MySQL, Galera, and cPanel/WHM; each tab fills in when the agent detects that stack on that server.
ServersCTL does not host traffic. It moves DNS, queues remote jobs, and calls APIs where configured.
Pool Presets
Server pools are created using a preset template in the UI. This chapter is for the Generic Linux Server Preset. For HAProxy Server Pools, see the HAProxy chapter.
What runs on a member (stack compatibility)
Do not assume one VM runs every stack. Common deployments:
Core terminology
Architecture
Failover health: missed heartbeat beyond failover delay (10–120 s). No HAProxy systemd check on hosting presets.
All Linux Servers should use the Generic Server Preset when adding a pool. Only ever select the HAProxy Preset if HAProxy is installed on your server.
Create a Generic Linux pool
Click Create Enrollment" and copy the install command.
Add further members to the Pool
How pool navigation works
Two layers on one screen
Protection and Managed DNS are pool settings. They float in the top tab bar above the member server tabs. They coordinate account replication, DNS cutover, and provider keys across the fleet — not operations on a single box.
Pool tab visibility (Generic Linux template)
Tab: Overview (first pool tab).
Purpose: Fleet-wide health — are agents reporting, are backups and jobs healthy across Linux servers?
What you see
Operator actions
Tab: Protection (pool tab bar).
When visible: 2+ pool members where the agent reports cPanel.
Purpose: Account-level warm standby — scheduled WHM backup → Secure Storage → restore on a standby server, with optional DNS cutover per protected account.
What you see
Operator workflow
1h … 1mo), DNS TTL, DNS provider (Cloudflare or WHM).
Replicate now / Replicate protected (Pro) — on-demand sync.
Account Cut DNS — per-account A record swing to standby (coordinates with Managed DNS).
After DNS cut: post-failover hook on standby.
Relationship to member tabs
Protection is pool-wide orchestration; member cPanel is per-server WHM operations.
Replication Transfer
Replication can take anywhere from a few minutes to several hours, depending on the size of the account.
The source agent will package the account and split it into multiple chunks, which are securely stored temporarily in D2 storage. Once all chunks have been uploaded, the UI instructs the receiving agent to download them and begin the restore process.
After the restore has completed successfully, all stored chunks are automatically removed from S3. As a guide, a 1.8GB backup typically takes around 5 minutes to replicate. Please take replication time into account when configuring your schedule. If the account is large, you may need to replicate once per day or every few days to avoid overlap and ensure the process completes cleanly.
Tab: Managed DNS (pool tab bar).
When visible (Generic Linux): 2+ cPanel-detected members
Purpose: Pool-level DNS catalogue for protected accounts and optional dynamic DNS — WHM vs Cloudflare zones, record health, sync, import. The API keys listed here are for DNS only. Do not use your production WHM API key here. You must use Cloudflare or a cPanel DNS Cluster API Key.
What you see
Operator actions
Protected DNS
The UI treats WHM servers as the source of truth and replicates changes to any linked DNS provider, as long as the account is marked as Managed. By default, the only record that will cut over automatically is the domain’s A record.
Multiple DNS Record Cut Over
You can configure the UI to cut over additional DNS records from the Protected Account DNS list. From here, you can specify A, AAAA, MX, and SRV records that should automatically cut to a standby server when the primary becomes unavailable.
Actions
Disable Managed DNS – When disabled, DNS records will not be updated or cut over during failover.
DNS Only – During cutover, Cloudflare proxying will be disabled (grey cloud), ensuring a direct DNS‑level switch without CDN caching or WAF interference.
Sync to Cloudflare – If you’ve added new DNS records in WHM’s DNS Manager, they will appear in the Protected DNS table and can be automatically pushed to Cloudflare.
Pool vs member DNS
Tab: Monitoring (pool tab bar).
Purpose: Pool-wide monitoring presets — distinct from per-member alerts on each server’s Monitoring tab.
Generic pool without Protection (0–1 cPanel members)
Infrastructure alerts section:
Generic pool with Protection (2+ cPanel members)
Protection DNS failover section (in addition to or instead of infrastructure, depending on layout):
Operator note
Configure per-member heartbeat and resource alerts on each server’s member Monitoring tab. Pool Monitoring is for fleet-level and protection DNS behavior.
Tab: Settings (pool tab bar).
Purpose: Pool identity, shared API credentials, danger zone.
What you see (Generic Linux)
What is NOT on the Generic pool Settings
HAProxy pools include Balancer failover on Settings.