Standardised virtual servers on BIG-IP — requested by your teams, approved by you.
SVX is a lightweight portal that lets application teams raise virtual-server requests against your naming standard, and lets network administrators review, approve and push them to BIG-IP — without ever handing out device access.
| Virtual server | Destination | Pool | Status |
|---|---|---|---|
| crm_443_vs | 10.20.5.10:443 | crm_8443_pool · 2/2 up | available |
| pool crm_8443_pool · monitor crm_https_monitor · node_10.20.9.11:8443 ● node_10.20.9.12:8443 ● | |||
| portal_443_vs | 10.20.5.11:443 | portal_80_pool · 1/2 up | degraded |
| ldap_636_vs | 10.20.5.20:636 | ldap_636_pool · 0/2 up | unavailable |
| api_443_vs | 10.20.5.12:443 | api_8080_pool · 3/3 up | available |
| crm_80_vs | 10.20.5.10:80 | redirect → 443 | available |
The Dashboard — every virtual server with its pool and nodes, live from the device. (Illustrative data.)
Virtual-server requests without the ticket ping-pong — or the device logins.
Most BIG-IP estates have the same problem: application teams need load-balanced services quickly, the network team is the bottleneck, and every “just give me access” request is a risk. SVX puts a standard, reviewable request in between.
A naming standard that enforces itself
Users type an application name. The virtual server, pool, nodes, profiles, persistence and health monitor are all named for them (crm_443_vs, crm_8443_pool, crm_http…) and validated server-side. No more free-text names.
Nothing reaches the device unreviewed
Requests go to a staging queue. An administrator sees the complete, resolved configuration — every object that will be created — and approves with a comment or rejects with a reason.
The audit trail writes itself
Who asked, who approved, what was pushed, when, and whether it worked — recorded against every job and in a dedicated audit log. The queue history downloads as CSV or JSON.
Everything a controlled BIG-IP change needs. Nothing it doesn’t.
Built around how network teams actually work: one change ticket, several virtual servers, a reviewer who needs to see exactly what will happen before it does.
Staging queue & approval
Build up several virtual servers, review them all, push in one go. Approvals require a comment; rejections require a reason the requester can read.
Standard naming, automatically
All object names derived from the application name, with live previews as you type. HTTP, TCP or UDP per virtual server, each with its own per-app profiles.
Full configuration preview
Each queue card shows the pool and members, monitor, every profile and persistence it will reference — badged will create, existing or built-in. Download the equivalent tmsh commands for review or manual application.
Clean deletes
Deleting a virtual server removes the pool, nodes, profiles and monitor the portal created for it — and keeps anything another virtual server still uses, telling you which one.
Live dashboard
Virtual server → pool → node health in one tree, connections and throughput, device CPU and memory, SSL certificate expiry. Search, status filters, Top 5 lists, CSV/JSON/print exports.
Multi-device & replication
Manage several BIG-IPs from one portal. Replicate an approved configuration to another device with new addresses in a couple of clicks.
Roles, two-factor, audit
Read-only, read/write and admin roles. TOTP two-factor authentication is required for anyone who can change a device. Two log streams, readable in the portal.
Safe by default
A failed create rolls itself back. Profile attachment is verified after every push. HTTPS out of the box, your own certificate the moment you drop it in.
How it works
Three steps, one service account, zero device logins for users.
Request
An application owner enters the application name, virtual server address and port, protocol, and pool members. Names are generated, the configuration is validated, and the request is staged — not applied.
Review
A network administrator opens the queue and sees the full resolved plan for each request. They approve with a change comment, reject with a reason, or replicate it to a second device.
Push
The portal applies the change over iControl REST using its own service account, verifies the result on the device, rolls back on failure, and writes the audit record. Undo is one click — it raises a reviewed deletion.
Built for change control
Auditors ask “who changed what, when, and why”. SVX answers all four without anyone keeping a spreadsheet.
- Every approval and every rejection carries a mandatory comment or reason
- Downloadable change history — CSV for the flat record, JSON with each full configuration
- Dedicated audit log, separate from application logs, with a request ID tying each push together
- Delete preview names exactly what will be removed, and what will be kept and why
- Undo a push by raising a reviewed deletion — nothing is ever silently reverted
tmshexport of any queued change for out-of-band review
Security that a network team will sign off
Designed for management networks: nothing leaves your network, nothing phones home, and the BIG-IP credentials never reach a browser.
One service account
The portal talks to BIG-IP with a single, least-privilege iControl REST account held on the server. Users authenticate to the portal, never to the device.
Two-factor authentication
TOTP with any authenticator app, with single-use backup codes. Required for every account that can change a device.
Hardened web layer
bcrypt-hashed local accounts, CSRF protection, login rate-limiting, strict security headers, HTTPS terminated in-process with your own certificate supported.
Server-side enforcement
Every name and address is validated on the server before it is embedded in a device call. Roles are enforced on every endpoint, not just hidden in the UI.
Audit log
A separate, rotated audit stream records logins, every queue decision, every object created or deleted on a device, and every denied request.
Air-gap friendly
No CDN calls, no telemetry, no licence server. Licence keys are signed files you copy onto the host. Updates arrive as checksummed bundles.
Installs in minutes. Upgrades in one command.
Delivered as a self-contained offline bundle with SHA-256 checksums. Choose the packaging that fits your host, then follow the setup guide.
Docker
A single container with Compose. ./install.sh loads the shipped image, asks for your BIG-IP details, generates the admin account and TLS certificate, and starts the portal.
Native (no Docker)
For hosts that can’t run containers — including unprivileged LXCs. Every Python dependency is vendored; the installer builds an isolated environment offline and installs a systemd service.
Upgrades with a safety net
./upgrade.sh --bundle verifies the bundle, snapshots your data, keeps a rollback image, upgrades, and confirms the new version booted. --rollback reverts.
Requirements
A Linux host with network reach to the BIG-IP management address, and a BIG-IP service account with iControl REST access to the target partition. Built against iControl REST; verified on BIG-IP 21.x.
Download SVX
The full product. The 30-day trial starts on first run, so there is no sign-up and no key to request. Pick the bundle that suits your host, copy it to the server, and run the installer.
Docker bundle
Recommended. A pre-built image, Compose file and installer, for any Linux host with Docker and Docker Compose.
tar xzf svx-docker.tar.gz
cd fp-f5-ss-*/
./install.sh
Download for Docker
Native bundle (no Docker)
For hosts that can’t run containers, such as unprivileged LXCs. Needs Linux x86_64 with Python 3.12; runs as a systemd service.
tar xzf svx-native.tar.gz
cd fp-f5-ss-*/
./install.sh
Download native
Setup and user guide · Every release lists SHA-256 checksums and release notes. All versions and checksums · Install and upgrade guide
Simple licensing
Try the full product free for 30 days — the trial starts automatically on first run. Then license it per portal installation.
30-day trial
- The complete product, no feature limits
- Starts automatically on first run
- No card, no sign-up, no internet access needed
- Keep your configuration when you license
Annual licence
- Licensed per portal installation
- 12-month signed licence key, installed offline
- All releases during the licence term
- Email support from the team that builds it
Questions we get asked
Does it need internet access?
No. The install bundle contains everything, there is no licence server, and the portal makes no outbound calls. It is built for F5 management networks where GitHub, PyPI and Docker Hub are blocked.
Can it change my BIG-IP outside the queue?
No. The only path to a device is a queued request that an administrator has approved. There is no “apply now” route, and deletions go through the same review.
What permissions does the service account need?
An iControl REST account that can read and write LTM objects (virtual servers, pools, nodes, profiles, monitors) in the partition you manage, plus read access to statistics and the certificate store for the dashboard. We supply a permissions guide with the bundle.
Can I use my own TLS certificate?
Yes. Drop your certificate and key into the certs/ directory and restart; it is used from then on and never overwritten. Out of the box the portal generates a self-signed certificate for the hostnames you give it.
What happens when the licence expires?
The portal shows a licence-expired page and stops serving requests until a new key is installed. Nothing on your BIG-IP is touched, and your queue, users and devices are kept.
Does any of my data leave my network?
No. The portal stores its state in local files on the host and talks only to the BIG-IP devices you configure.
Is this an F5 product?
No. SVX is developed and supported by DDR Computing Ltd and uses the standard BIG-IP iControl REST API. F5 and BIG-IP are trademarks of F5, Inc.; DDR Computing Ltd is not affiliated with or endorsed by F5, Inc.
Ready to standardise your BIG-IP changes?
Tell us how many devices you run and whether you’d prefer Docker or a native install, and we’ll send the trial bundle and a short guide to get it running.