PCI-DSS & GDPR

Every feature, on every plan.

Every feature, on every plan, at every size. No add-ons, no feature tiers, no upgrade to unlock. Choose the resources that fit the workload, not the features.

High severity e2ac2eee
2026-08-22

Remove expired generated sales rules

getActiveAttributes · request hot path

UPDATE salesrule SET is_active = 0 WHERE is_active = 1 AND to_date < CURDATE()

expired rules still active · refilled by the generator

Fix Deactivate expired cart price rules in one bounded batch, then refresh the salesrule indexes.

Run by the platform. Marked applied.

runtime 183.6 ms 0.94 ms

A recommendation delivered to a Magento store on synsmarts during its migration review. Numbers as measured. Store name withheld.

Built For PHP

Hosted as

99.99% Uptime. Built to stay up, and watched all day by people who can act.
15 min A real person on the panic call, 24/7, on every plan.
30 days Automatic backups. Rewind to any minute.
PCI-DSS + GDPR Documented, and your assessor can ask for it.

Everything a store needs, already in the box.

The stack every open-source store expects, on every plan, with no add-on and no upgrade to unlock it. This is the host. What comes after is what makes it different.

See what the stack costs
Every store gets its own home with the full stack, a burst pool that scales out under load, and its own cron pool. Nothing is shared with another store.

Runtime

  • PHP 8.2, 8.3, or 8.4

    Pinned per app, with OPcache and the extensions you pick.

  • MySQL or MariaDB

    Your own database, backed up hourly, restorable to the minute.

  • Redis

    Sessions and cache on their own instance, kept across every deploy.

  • Varnish

    Full-page cache for Magento and Shopware, one toggle.

  • OpenSearch

    Your own cluster, heap sized from the memory you reserve.

  • RabbitMQ

    Your own broker, enabled with one setting, queues and consumers ready.

  • Nginx

    Your own vhost, with server and location snippets you write.

Manage

  • A dashboard for every store

    Every setting for every store, one login, no ticket to change any of it.

  • One config file, versioned

    Your store's runtime declared in a file you own, with every change tracked.

  • Every service setting from a catalog

    PHP, nginx, MySQL, Redis, OpenSearch, RabbitMQ, Varnish. Vetted values, applied safely.

  • API, CLI, and MCP

    Every operation reachable from all three, including your own coding agents.

  • Environment variables and secrets

    Declared in the config, delivered to the store, never in the repo.

  • Deploy hooks and build steps

    Run what you need before, during, and after every release.

  • Bring your own Dockerfile

    Extend the base image when the store needs more than the defaults.

  • Maintenance page

    Your own page, on or off from the config file or the dashboard.

Deploy and scale

  • Git push, CLI, or API to deploy

    From GitHub through our app, or from your workspace with Deployer or rsync.

  • Blue/green with automatic revert

    A health probe before traffic moves, and the old release stays live if it fails.

  • Live workspace preview

    Edit over SSH or rsync and see it on a preview URL. Production changes only when you deploy.

  • Cron you schedule

    Your schedules, with run history for every job.

  • Autoscaling: off, capped, or unlimited

    Per store, with the first 10% of the month up to 3x free.

  • Storage that grows on its own

    25 GiB at a time, with an alert at 80% and a ceiling you set.

Access

  • SSH and SFTP with your keys

    Your keys, your own shell pod, session recorded for your PCI evidence.

  • A real toolbelt

    composer, wp-cli, git, node, rsync, mysql and redis clients, ripgrep, jq, tmux, and the synsmarts CLI to deploy or restore from the shell.

  • Direct MySQL over an SSH tunnel

    Your own client on port 3306 through the bastion, no shared password.

  • Team roles

    Owner, admin, developer, billing, each with only the access it needs.

  • Two-factor login

    TOTP with recovery codes, available on every account.

  • Twelve months of team activity

    Every login, change, and deploy, who did it and when.

  • Log streams from the shell

    PHP-FPM slow, MySQL slow, and your app's own logs, live.

Edge and network

  • SSL on any domain

    Issued and renewed for you, on domains you host with us or keep elsewhere.

  • DNS in the dashboard

    Host the zone with us, delegate it, or keep your own.

  • Multi-store domains

    Magento store codes routed per domain.

  • Cloudflare CDN and DDoS

    Every store fronted by the edge, with the origin locked to it.

  • Per-store firewall and WAF

    IP allow and deny, rate limits, path rules, and OWASP rules tuned to your app.

  • Email from your own domain

    Transactional mail with DKIM, reputation isolated from every other store.

The difference

What happens when something breaks.

Before anything changes, synsmarts captures the evidence, identifies the root cause, and recommends the best path forward.

Traditional hosting

  • Restart first. Investigate later.
  • "The server is fine. The problem is your code."
  • "Here is the graph. You figure out why."
  • Essential tools often locked behind higher plans.

synsmarts

  • Capture every request before taking action.
  • Find the root cause, to the file and line.
  • Get a written diagnosis with the patch for your code.
  • Every essential feature included on every plan.

The whole picture, so nobody has to guess.

One slow checkout, one trace: edge, cache, PHP, the query, and the line of code, timed end to end. Metrics, logs, and events are watched all day for every store. A trace is captured the moment you or the platform asks for one, complete, never sampled.

synsmarts mcp · query_traces

$ query_traces --app store --since 14d --sort slowest
→ edge → cache → PHP-FPM → MySQL, one trace
/checkout 14.2s
/cart 2.1s
/api/products 118ms

Traces

On demand, browser to database, never sampled.

Metrics

Per store, all day.

Logs

PHP, slow query, slow request, and your app's own.

Events

Deploys, crons, restarts, scaling.

Configs and code

Read alongside the trace.

Watched all the time

Detection runs on every store around the clock. When it fires, a diagnosis starts on its own and you hear about it with the cause, not a graph.

Twelve months of audit history

Every login, change, and access, kept for a year, so when your assessor asks for evidence you already have it.

The fix arrives with the diagnosis.

Every issue follows a clear path, from detection to resolution, so every fix starts with evidence instead of guesswork.

  1. Detect
  2. Diagnose
  3. Report
  4. Recommend
  5. Contain
High severity 5ea5014e
incident · 2026-08-22

Reduce PHP-FPM pool concurrency

php-fpm · pm.max_children

12 workers exhausted safe process headroom during the incident.

Fix

Set pm.max_children to 9. Prevents recurrence while keeping bounded request concurrency.

Apply settings change

A report you can act on

Root cause to the file and line, the evidence, and the patch for your code. Everything needed to fix the issue.

Your code, indexed next to the trace

Your store's code is indexed alongside its traces, so the report names the exact code behind every issue.

Caught before it ships

Every release is verified before it goes live and rolls back on its own if a gate fails. Identify issues before they reach production.

The platform fixes what you let it fix.

Before anything changes, synsmarts captures the evidence, finds the root cause, and recommends the safest next step. A 3am incident is handled before you wake up, and you get the answer, not a shrug.

YOUR APPLICATION

You're in control.

We explain what happened and recommend the next step. Nothing happens to your store until you say so. No "we killed your pod" tickets, ever.

  • Restart services

    Only after the cause is known, never as a guess

  • Resolve stuck database queries

    The one query that's stuck, not the whole store

  • Pause background jobs

    Imports and indexing wait. Checkout doesn't.

  • Clear cached data

    Only the stale pages, so the rest stay fast

  • Security monitoring

    Threat detection can't be switched off, on any plan

    ALWAYS ON

Turn an action on and the platform handles it without you.

THE PLATFORM UNDERNEATH

Built to recover safely.

synsmarts takes the safest action first, keeping infrastructure healthy while protecting uptime. The evidence is captured before anything is restarted.

  1. 01

    Stop the affected process

    Resolve the issue without impacting everything else. One stuck thing is stopped, the rest of the store keeps selling.

  2. 02

    Redirect traffic

    Keep requests flowing while the platform recovers. Shoppers never see the repair.

  3. 03

    Replace unhealthy services

    Bring healthy resources online before retiring old ones. The new one is working before the old one goes.

  4. 04

    Restart only when needed

    A restart is always the last step, never the first.

And when the store is healthy, you build.

The same platform that finds the fault is the one your team ships on. Here is what that looks like day to day.

Get a Free Diagnostic

Your coding agent gets the same access we have.

Connect AI coding assistants to synsmarts and manage deployments, diagnostics, and more from one place. Bring your own AI and it works alongside ours, so your store ships at the speed your tools can move.

API, CLI, and MCP reference

Every action from the editor

Deploy, restart, restore, and more. Your agent runs the store without leaving the editor.

Logs, traces, and metrics on demand

Access logs, traces, and metrics instantly. Your agent starts from the answer instead of searching for it.

Claude Code, Cursor, and the rest

Connect with Claude Code, Cursor, and more. The same tools our own engineers use to run the platform.

synsmarts mcp

$ deploy --tenant acme --sha 9f3c1d2
✓ blue/green gated on health + canary

$ query_traces --endpoint /checkout
→ slowest span: cart_items · 14.2s
~35 tools

A bad deploy never reaches a customer.

Deploy from GitHub or SSH with confidence. Every release is verified before it goes live, and if something goes wrong, synsmarts automatically rolls back without disrupting users. No deploy windows. No held breath.

How the move works

EVERY DEPLOY

Blue, live 100%

Shoppers already in checkout finish their order here.

health + canary atomic switch
Green, new build 0%

Pre-warmed. If anything looks wrong, the old version stays live.

Keep users online during every deployment. No maintenance windows, no interrupted sessions, and no dropped requests.

Every change is tried on a copy of your database first.

THROWAWAY CLONE

clone_acme_9f3c1d2

4 statements · 0 rows rewritten

dry-run passed · 1.8 s

Clone. Test. Remove. Every change is tried against a copy of your real database first, so "works on staging, dies in production" ends here.

  • CODE-ONLY Go straight to deployment.
  • ADDITIVE Applied while traffic keeps flowing.
  • BREAKING Requires approval and a backup first.
  • CAN'T CLASSIFY Handled as breaking.

Fast every day, faster on sale days.

Autoscaling

More cores the moment the sale needs them, on their own. Up to 3x is free for the first 10% of the month, or set a hard cap.

Cloudflare edge

Security and speed from the edge. A shopper in Sydney loads the store as fast as one in Chicago.

OPcache + FastCGI

Ready from the start. No warm-up after a deploy.

Redis object cache

Survives every deployment. Sessions and carts stay intact.

S3 + CDN media

Fast media delivery. Product images load from the nearest city, not from your server.

Undo it yourself, to the minute, with no ticket.

Recover quickly with automatic backups, point-in-time restore, and secure copies, all included as standard. Break the store at 3am, pick the minute before, and undo it yourself. No ticket, no waiting for someone to find the file.

Restore is part of every migration too
PITR · ANY TIMESTAMP
XtraBackup · hourly
full · 30 days
Point-in-time recovery
any minute · 14 days
EBS snapshots · hourly
volume · 7 days
−30 d−14 d−7 dnow
30 days Of hourly copies kept
14 days You can rewind to any minute
Always on Nothing to schedule, nothing to remember
Yours to keep Download any backup, any time

Your assessor asks. We already have it.

The documents, ready

Attestation of Compliance, shared responsibility matrix, penetration test summary, and the DPA with Standard Contractual Clauses. One email to security@synsmarts.ai.

The evidence, kept

Twelve months of audit log, every deploy and access recorded, and the tenant isolation test results your assessor asks for under Appendix A1.

The security page, with the documents
PCI-DSS SAQ DGDPR with DPAIsolation tested

The edge Cloudflare absorbs DDoS and serves the cache. The origin answers only requests that came through it.

Your store's own firewall OWASP rules tuned to your app, with path rules you set. This is the boundary your assessor audits.

Your data

Its own database, its own encryption key, a copy in a second region. No other store can reach it.

Isolated per store

Everything you need to build anything you want.

Start with a free diagnostic and see what the platform finds on your own store.

$187.50 / mo starting

Grow when you're ready.