NGINX support

NGINX Support for Open Source, NGINX Plus and Ingress Estates

Last reviewed by Tyler Eastridge

AceMQ supports NGINX 24/7 across Americas, EMEA and APAC: open source mainline and stable builds, NGINX Plus configurations, reverse proxies, load balancers and Kubernetes Ingress. Most calls start with one of three problems: 502 or 504 errors from a proxied service, "worker_connections are not enough" under load, or TLS handshakes failing after a certificate change.

AceMQ provides independent 24/7 support for NGINX open source, NGINX Plus configurations and Kubernetes ingress layers: an AceMQ engineer and a partner NGINX specialist on a live bridge for a P1 (response targets from 15 minutes, by plan), a written root-cause analysis after each one, and version planning against nginx.org. NGINX open source has no vendor SLA, and F5 sells its own NGINX Plus subscriptions, so AceMQ is the escalation path for the estate you run, not a replacement for an F5 entitlement. AceMQ does not ship its own NGINX builds.

Support quotes answered within 1 hour, by a senior engineer.

Planning a build, upgrade or migration instead? See AceMQ NGINX consulting and services.

160+
Customers supported
10+
Years in production support
40+
Sev 1 incidents handled a year
24/7
Americas, EMEA and APAC
1hour
Quote reply

Customers

  • Microsoft
  • Henry Schein
  • Enbridge
  • Gojek

Partners

  • Broadcom
  • Redis
  • VMware Tanzu
  • OSSeva

Incident Triage

Problems we fix every week on NGINX

The proxy faults we get called on, each with the log line or directive we look at before suggesting a change.

Common NGINX production problems and how AceMQ resolves them
Symptom you're seeingWhat we check first
502 and 504 errors after a deployA 502 means NGINX could not get a valid response from the upstream: connection refused, reset, or a bad response. A 504 means it waited past proxy_read_timeout (default 60 seconds) or proxy_connect_timeout. The log line says "connect() failed", "upstream prematurely closed connection" or "upstream timed out". We check the backend's readiness during rollout, keepalive settings (keepalive in the upstream block, proxy_http_version 1.1 and an empty Connection header), and whether proxy_next_upstream retries a non-idempotent request, which can double-submit a payment.
"worker_connections are not enough"Each proxied request uses two connections, one to the client and one to the upstream. The ceiling is worker_processes times worker_connections, bounded by the file descriptor limit. We check worker_rlimit_nofile, the systemd or container LimitNOFILE, and the kernel's net.core.somaxconn and port range, and read stub_status (active, reading, writing, waiting) to see where connections sit.
A reload "succeeds" but old workers stay for hoursAfter nginx -s reload, old workers finish their connections and exit. Long-lived WebSockets, streaming responses and slow clients hold them. The process list shows "worker process is shutting down". We set worker_shutdown_timeout, check what holds the connections, and confirm a reload does not double memory use on a busy host.
Errors continue after a backend changed addressNGINX resolves hostnames in proxy_pass once at start or reload unless a variable is used together with a resolver. A backend behind a load balancer or a Kubernetes Service whose IP changed keeps receiving traffic at the old address until the next reload. We use resolver with valid= and variable-based proxy_pass where it is right, or upstream resolution on NGINX Plus.
TLS handshake failures after a certificate changeMissing intermediate certificates in the chain served by NGINX, a certificate and key mismatch, ssl_protocols or ssl_ciphers that old clients cannot meet, SNI and the wrong server_name block, or an OCSP stapling setting. We test with openssl s_client -connect host:443 -servername host and check the chain, then review the protocol and cipher set against your client population.
A cache stampede hits the originCached pages that expire together send a burst of identical requests. We use proxy_cache_lock, proxy_cache_use_stale updating error timeout, proxy_cache_background_update and sensible inactive and max_size values, and verify the cache key is not too broad or too narrow.
Redirect loops, wrong client IPs and headersBehind a load balancer, real_ip_header, set_real_ip_from and X-Forwarded-Proto handling decide whether the application sees the right client and scheme. A mismatch causes login loops and wrong rate limiting. We trace the header path from edge to application.
Ingress changes are not appliedOn Kubernetes, the Ingress controller generates the NGINX configuration. A rejected annotation, a failed admission webhook or a configuration reload failure leaves the old config running. The community Ingress NGINX controller was retired by the Kubernetes project in March 2026, so a configuration problem on it will not get an upstream fix (Kubernetes blog, 11 November 2025). We help you read the controller logs, and plan the move to Gateway API or another controller.
Security patches and version decisionsNGINX release notes carry security fixes. The nginx.org change log (read 17 November 2026) lists 1.31.7 (14 October 2026) with a fix for a buffer overflow in a worker process when alias is used and the request URI was changed by rewrite (CVE-2026-15176), and 1.31.6 (15 September 2026) with a heap buffer overflow when using HTTP/3 with OpenSSL 3.5.0 and earlier (CVE-2026-90439). We check whether your configuration uses the affected directives or modules before the change window, and plan the upgrade.

Each closed P1 names the directive, certificate or upstream at fault and the config change that fixed it.

Edge proxy throwing 502s on a live service? Request a quote and we reply inside an hour.

Contact Us

Deployment coverage

Which NGINX deployments does AceMQ support, and who covers what?

DeploymentWhat AceMQ coversWhat the vendor coversExcluded
NGINX open source, mainline and stableIncident response, configuration review, tuning, TLS, upgrades and security patch planningThe open-source build carries no vendor support contractWe cannot patch the NGINX source for you
Reverse proxy, load balancer, TCP and UDP proxyUpstream design, health, timeouts, keepalive, caching, rate limitsNot applicableYour backend application's code
NGINX Plus configurationsConfiguration, behaviour and incident help on your NGINX Plus estateF5 provides NGINX Plus under its subscription and supports the productThe F5 subscription stays with F5; we do not resell or replace it
F5 NGINX Ingress ControllerAnnotations, TLS, routing, upgrade planningF5 for the controller productProduct defects go to F5
Community Ingress NGINX controller (retired)Reading controller behaviour, stabilising what runs now, planning the move to Gateway API or another controllerThe Kubernetes project retired it in March 2026 and does not maintain itFixes to its code
Virtual machines and bare metalOS, kernel, descriptors, TLS and logsOS vendorHardware faults
Kubernetes and OpenShiftThe Ingress layer; see also Kubernetes supportThe platform vendorCluster control-plane faults
Air-gapped and regulated hostsSame scope inside your controlsNot applicableAnything needing outbound access

Support options

Who provides enterprise support for NGINX? F5 and independent support compared

NGINX open source has no vendor support contract. F5, which owns NGINX, sells NGINX Plus and related products with commercial support. Independent support from AceMQ covers open source NGINX and the Kubernetes Ingress layer around it, and can assist on NGINX Plus estates without replacing the F5 subscription.

F5 (NGINX Plus and related products)
AceMQ independent support
What it coversNGINX Plus, F5 NGINX Ingress Controller and other F5 NGINX products under subscription
What it coversOpen source NGINX (mainline and stable), NGINX Plus configurations, Ingress layers, and the proxy, TLS and upstream behaviour around them
Open source NGINXNot part of a Plus subscription
Open source NGINXCovered
Product defectsF5 fixes and ships its own product
Product defectsWe diagnose, gather evidence and work the escalation with F5 for Plus-specific issues
Root cause analysisPer F5's terms
Root cause analysisWritten root cause analysis on every P1
PriceSet by F5
Price$24,000 to $72,000 a year (USD), set by SLA, scope and deployment. Quote reply within 1 hour

Dated source: F5 NGINX Plus product page, f5.com/products/nginx/plus, read 17 November 2026 (it does not state support hours or response times). Short note: if you need NGINX Plus features, buy the subscription from F5. If what you need is an engineer when something breaks, in any NGINX build, that is a separate decision. Many estates do both.

What's Included

Everything in your NGINX support contract

Open-source NGINX configurations, reverse proxies and load balancers, TLS, caching and the Kubernetes Ingress layer, under one agreement.

  • 01

    Emergency Incident Response

    Proxy returning errors, certificates failing, an ingress layer not routing or a fleet out of connections. With response targets from 15 minutes, by plan, an AceMQ engineer and a partner NGINX specialist join a live bridge, 24/7, to diagnose, not to acknowledge a ticket.

  • 02

    Root Cause Analysis

    Every P1 closes with a written RCA: what failed, why, the change that fixed it and the setting that prevents a repeat. Delivered as standard, not on request.

  • 03

    Version and Upgrade Planning

    Where each proxy sits against the mainline and stable lines on nginx.org, and the path to the line you choose, rehearsed on a copy of the configuration. The detailed upgrade and migration work is on the NGINX services page.

  • 04

    TLS and Exposure Review

    Certificates and renewal, protocol settings, default servers, headers and which modules are compiled in, read against how the proxy is reached from outside.

  • 05

    Connection and Performance Diagnosis

    Worker, descriptor and per-upstream limits, buffers, keepalive and cache behavior traced from logs and metrics rather than guessed at.

  • 06

    Kubernetes Ingress Layer

    Ingress NGINX, F5's NGINX Ingress Controller and Gateway API controllers running in your clusters: reload failures, route conflicts and certificate handling, with the exit from Ingress NGINX planned alongside.

Why AceMQ

What teams value about AceMQ for NGINX

  • Engineers Who Know Your Environment

    Tickets go to senior engineers briefed on your server blocks, upstreams and certificate setup, so a P1 call starts with a hypothesis rather than an interview. The proposal states who does the work.

  • No Help-Desk Layer

    You reach a senior engineer directly by phone, email or Slack. No queue collecting details to relay, and no approval between you and someone who can read an error log and act.

  • Honest About What We Cannot Patch

    AceMQ does not ship NGINX builds or backport fixes. Fixes come from nginx.org releases and, for NGINX Plus, from F5. For a proxy on a version without a fix, we show what is exposed and plan the upgrade.

  • Diagnostics Before Changes

    Logs, metrics and the effective configuration first, then a change with a rollback point. No directive is changed to see what happens in production.

  • Cover Across Time Zones

    AceMQ staffs 24/7 across the Americas, EMEA and APAC, so your 3am alert reaches someone in working hours.

  • One Contract Across the Platform

    NGINX rarely runs alone. One AceMQ contract and SLA can also cover Kubernetes, Kong, the monitoring stack and the databases behind the proxy, with multi-technology discounts.

NGINX estates we support

Single hosts, container images, Kubernetes controllers and cloud instances. NGINX Plus subscriptions remain with F5.

  • NGINX open source, mainline and stable lines
  • HTTP servers, reverse proxies, load balancers and TCP and UDP proxy servers
  • NGINX Plus configurations (the F5 subscription stays with F5)
  • Community Ingress NGINX controller (retired by Kubernetes)
  • F5 NGINX Ingress Controller
  • Virtual machines and bare metal
  • Kubernetes, including EKS, AKS and GKE
  • Containers built from official or custom images
  • AWS, Azure and Google Cloud instances you run
  • Hybrid estates

Version coverage

Versions we support

Versions we support
LineSupported by AceMQ
NGINX 1.31 (mainline)Yes
NGINX 1.30 (stable)Yes
NGINX Plus releasesConfiguration and behaviour; the subscription stays with F5
F5 NGINX Ingress ControllerYes
Community Ingress NGINX controllerOperation and migration planning (see problem 8)

Source: nginx.org download page (as used by the existing version table, read 4 November 2026); nginx.org change log, read 17 November 2026, latest mainline listed there 1.31.8 (21 October 2026).

In practice

The five log fields and checks we ask for first

A stable NGINX estate is mostly a matter of what you can see. When a customer calls with an outage, the questions below decide how quickly we can find the cause, and we put them in place during onboarding.

An access log format that carries the upstream. Add $upstream_addr, $upstream_status, $upstream_response_time, $request_time and $upstream_connect_time to the log format. With those, a slow request splits into time spent connecting, time the backend took and time spent sending the response to a slow client. Without them, every slow request looks like an NGINX problem.

Error log at the right level. The warn level shows upstream timeouts and connection failures. info is noisy. debug needs a debug-enabled binary and should be turned on for a single client address only (debug_connection).

A stub_status endpoint, or NGINX Plus API metrics. Active connections, reading, writing and waiting show whether you are near the connection ceiling. Scraping them into your monitoring gives a trend, not only a snapshot.

A configuration test in the deploy path. nginx -t before every reload catches syntax errors, but not logic errors such as a location block shadowing another. Reviewing the output of nginx -T, which prints the full effective configuration including includes, is how we find those. We keep that output in version control after each change so a diff shows what moved.

A rollback you have tested. A reload that fails leaves the old configuration running. A reload that succeeds with a bad configuration serves errors. Keeping the last known good configuration, and testing the rollback, turns a long outage into a short one.

We also check three platform settings that cause intermittent trouble: the number of worker processes against cores (worker_processes auto in most cases), the keepalive timeout to clients and upstreams so connections are reused without hoarding, and the TLS session cache so repeat handshakes are cheap. None is exciting. Each has turned up in a P1 root cause analysis.

Who stands behind NGINX

Open source, F5 and AceMQ: who does what

Three parties are involved when NGINX misbehaves, and a page that blurs them sends buyers into the wrong decision.

  • What AceMQ provides

    Engineers who run, troubleshoot and upgrade NGINX, on-call response, root-cause analysis, version planning and performance diagnosis, under one AceMQ contract and SLA.

  • What stays with nginx.org

    The open source releases. Fixes and security advisories come from the mainline and stable lines, and AceMQ does not fork, rebuild or backport them.

  • What stays with F5

    NGINX Plus, F5's NGINX Ingress Controller and their support entitlements. AceMQ is independent of F5 and does not give F5 licensing advice.

  • If you are on the retired controller

    The Kubernetes project archived Ingress NGINX and will publish no more fixes. AceMQ supports it while the exit on the NGINX services page is planned and executed.

  • Who this is not for

    Teams that need NGINX Plus features or F5's own support entitlement should buy those from F5. A single reverse proxy that rarely changes and has no on-call duty may not need a 24/7 contract.

Escalation Path

Your first hour of a NGINX outage

When the edge proxy returns 502 for a live service, every user sees it, so that incident gets the quickest clock.

  1. From 15 min (by plan)first responseEmergencyThe edge proxy returning 502 or 504 for production, worker connections exhausted at peak, or a certificate change breaking every client
  2. 1 hourfirst responseCriticalOld workers piling up after reloads and eating memory, Ingress changes not applying on a production cluster
  3. 4 hoursfirst responseHighA cache stampede on the origin after a content change, wrong client IPs behind a load balancer, TLS failures for old clients
  4. Next Dayfirst responseStandardMainline versus stable choice, upstream keepalive tuning, Gateway API migration planning

Timers depend on plan. A proxy failing for production traffic starts the fastest one, a mainline-or-stable question the slowest.

Need an engineer who already knows your upstreams and TLS setup? Request a quote or write to us.

Contact Us

Pricing

What does AceMQ NGINX support cost?

NGINX support typically costs $24,000–72,000 a year. The figure tracks how many proxy instances and environments you run, whether Ingress NGINX or another controller is involved, TLS and caching complexity, and the response level you need.

Prices are in US dollars (USD). Response targets start at 15 minutes by plan, 24/7, and quote requests are answered within 1 hour. Covering your Tomcat, Kong or other application tier in the same contract earns a multi-technology discount.

Typical engagements

NGINX faults we have diagnosed

Proxy, TLS and Ingress failures that we have traced to a directive, a backend or a controller setting.

  • Assessment

    Ingress NGINX to Gateway API Migration Assessment

    Platform Engineering Team

    A typical engagement for a platform team that still runs the retired Ingress NGINX controller: every Ingress resource, annotation and snippet inventoried and mapped route by route to Gateway API resources, with a parallel-run plan.

    NGINXKubernetesGateway APIIngress NGINX
  • From our platform practice

    Java Estate Inventory and OpenJDK Distribution Selection

    Mid-Sized Distribution Company

    A typical engagement for a company that wants to leave an Oracle Java SE subscription or avoid one: inventory every Java runtime, compare OpenJDK distributions against the estate, and write the migration plan.

    JavaOpenJDKDockerKubernetes
All use cases

Support options

Four tiers of NGINX support.

AceMQ offers NGINX support in four tiers, each handing more of the day-to-day work to AceMQ: 24/7 support and emergency response for proxies and ingress layers your team runs, proactive reviews that find certificate, connection-limit and version problems first, managed NGINX on your own infrastructure, and managed NGINX on AceMQ infrastructure. Open source mainline and stable lines, NGINX Plus configurations and Kubernetes ingress controllers are covered by AceMQ engineers working with partner NGINX specialists. Fixes still come from nginx.org and F5.

  • Tier 1

    NGINX 24/7 Support & Incident Management

    You manage NGINX, we take the hard calls

    Best for: Operations teams who own their configs and want a specialist for 502s, TLS breakage and reload trouble.

  • Tier 2

    NGINX 24/7 Support with Proactive Monitoring

    Config, TLS and version reviews on a schedule

    Best for: Teams who want certificate chains, connection limits and patch levels checked before a peak or an expiry.

  • Tier 3

    Managed NGINX on Your Infrastructure

    We operate proxies and Ingress on your hosts

    Best for: Organizations keeping every proxy node and cluster inside their own network, with a thin platform team.

  • Tier 4

    Fully Managed NGINX on AceMQ Infrastructure

    We run your edge proxy layer as a service

    Best for: Product teams who need reverse proxying and load balancing working without owning NGINX operations.

FAQ

NGINX support questions

Still have questions about NGINX? Email an Expert

Who provides enterprise support for NGINX?

F5 supports NGINX Plus under its subscription, and open-source NGINX has no vendor contract. AceMQ covers the estate you run, 24/7, with response from 15 minutes by plan. A Sev 1 with 502s on a live service means an engineer reading the error log and upstream timings first.

Is there commercial support for NGINX open source?

Yes, from independent providers. AceMQ offers incident response, config review, tuning and patch planning across mainline and stable builds. You receive a written root cause after each P1, and F5 stays the contact for Plus features.

Can AceMQ replace our NGINX Plus subscription?

No. NGINX Plus features and F5's support for the product come from F5. We support the estate around it and can work an escalation with F5 on your behalf, but we do not resell or replace the subscription.

Why do we get 502 or 504 errors from NGINX?

A 502 means no valid upstream response: refused, reset or malformed. A 504 means a timeout. The error log names which. We tie each to a backend using $upstream_addr and $upstream_response_time, then fix readiness, keepalive, timeouts or retry settings.

What does "worker_connections are not enough" mean?

NGINX has hit its connection ceiling: worker_processes times worker_connections, limited by the file descriptor limit. Proxying uses two connections per request. We raise the limits consistently across NGINX, systemd and the kernel and check what holds connections.

Why do old NGINX workers stay after a reload?

They are finishing long-lived connections such as WebSockets and slow downloads. worker_shutdown_timeout caps this. We find what holds them and decide whether the cap is safe.

Does AceMQ support the Ingress NGINX controller Kubernetes retired?

We support running it and planning the move. The Kubernetes project retired it in March 2026 and does not maintain it, so we stabilise what runs now and help you migrate to Gateway API or another controller.

Should we run mainline or stable?

Stable receives only critical fixes; mainline carries new features and fixes first. Most production estates choose stable unless they need a mainline feature, and plan a regular move to the next stable line. We decide with you from the modules and features you use.

How do we fix NGINX TLS handshake errors?

Test with openssl s_client, check the served chain, certificate and key match, SNI, and protocol and cipher settings against your clients. Most failures after a certificate change are an incomplete chain.

What does 24/7 NGINX support cover with AceMQ?

Round-the-clock incident response, written root cause on every P1, configuration reviews, upgrade and patch planning, and the Kubernetes Ingress layer. A typical Sev 1 is a certificate change that leaves clients unable to complete a TLS handshake.

Can AceMQ help with an Ingress NGINX move on a one-off basis?

Yes. An Ingress NGINX to Gateway API plan, a version upgrade or a TLS review can each be a single engagement. A yearly agreement adds 24/7 cover and named engineers, and some teams move onto it after the migration so the same people stay on call.

What drives the price of NGINX support?

Annual contracts run $24,000 to $72,000 (USD). We look at instance and environment counts, open source or NGINX Plus estates, Ingress use on Kubernetes, how much TLS and caching configuration you carry, and the response level. Send a summary of the layout and a quote follows within 1 hour.

Let's talk

Get 24/7 Support for Your NGINX Estate

Tell us your NGINX builds and where they run. We reply to quote requests within 1 hour.

A senior NGINX engineer, not a sales queueQuote requests answered within 1 hourAd hoc engagements or an annual partnership

Production incident right now?Email a NGINX Architect