Website or application unreachable
HTTP errors, blank page, timeout, connection refused, or an application that no longer responds normally.
Linux server diagnostics and intervention
Linux Server Troubleshooting and Diagnostics
Your website, application or server is no longer behaving as expected? Axanti can review your Linux environment, identify the cause of the problem and work on the server components involved.
We can work on a server hosted with Axanti or with another provider, when the necessary access is available.
Symptoms
Is your server showing one of these symptoms?
These are starting points, not diagnoses: the same symptom can have several causes depending on the environment.
500, 502, 503 or 504 errors
The code points at a problem, not at its cause: it can come from the web server, PHP-FPM, a proxy, the application or the available resources.
Slow or overloaded server
High CPU, not enough memory, stuck processes, slow queries, missing resources or a configuration that does not match the real load.
Full disk or not enough space
Logs, backups, databases, caches or application files that end up filling the available space.
A service that will not start
Nginx, Apache, PHP-FPM, MariaDB or MySQL, or another service the web environment depends on.
Problem after an update
A PHP version change, a system package, DirectAdmin or CustomBuild, a library or a configuration change that caused a regression.
Certificate, DNS or connectivity
An expired or misconfigured certificate, an incorrect DNS record, a resolution, redirect or service access problem.
Email on the server side
SMTP, IMAP, Postfix, Dovecot or server-side configuration, when email is part of the environment involved.
Scope of the analysis
An analysis across the whole server stack
An incident is rarely understood by looking at a single layer. What gets examined depends on the symptom described and on the environment in place.
Linux and the system
- ✓systemd services
- ✓CPU and memory
- ✓Storage and disk space
- ✓Permissions and processes
- ✓System logs
- ✓Packages and dependencies
Nginx and Apache
- ✓Virtual hosts
- ✓Reverse proxy
- ✓Configuration errors
- ✓Communication with PHP-FPM
- ✓Certificates
- ✓Redirects
PHP and PHP-FPM
- ✓Stopped services
- ✓Sockets and pools
- ✓PHP versions
- ✓Extensions
- ✓Resource limits
- ✓Errors after a version change
MariaDB and MySQL
- ✓Service unavailable
- ✓Connection problems
- ✓Load and slow queries
- ✓Configuration
- ✓Errors affecting the applications
DirectAdmin and cPanel
- ✓Server configuration
- ✓Control panel services
- ✓PHP versions and web stack
- ✓DNS and certificates
- ✓Problems tied to an update
DNS, SSL and email
- ✓Resolution and records
- ✓Certificates and renewals
- ✓Server-side mail services
- ✓Reviewed according to the problem at hand
Method
From the symptom to the cause
A diagnosis does not automatically lead to a complete repair. Depending on what is found, the next step may belong to a developer, a restore, a migration, an infrastructure change, the hosting provider, or a deeper security review.
01
Understand what changed
The symptoms, when the problem started, any recent changes or updates, and which services are affected.
02
Review the state of the server
Services, resources, system and application logs, and the configuration relevant to the described symptom.
03
Identify the likely cause
Separate the symptom from its real cause before changing anything in the configuration.
04
Fix within the agreed scope
Act on the server components involved once the cause and the scope of the work are clear enough.
05
Check that it works
Test the services involved after the work and flag whatever still needs watching.
No contract needed
You can call on us for a one-time intervention
You do not need a managed services contract with Axanti to have a server problem analyzed.
Depending on the situation and the access available, Axanti can work on the existing environment, run the diagnosis and carry out the agreed fixes. If you then want ongoing support, that is discussed separately and is never a condition.
Describe the problem
The quote form also works for an intervention request: describe the symptom and the context, and we come back to you with how the analysis would be approached.
Existing server
Your server can stay with your current provider
Troubleshooting does not automatically require a migration to Axanti. We can work on a VPS or a dedicated server hosted with another provider when the access and the environment allow it.
If the incident reveals a broader need — maintenance, updates, follow-up over time — recurring management is discussed separately, once the immediate problem is settled.
- ✓VPS or dedicated server with another provider
- ✓Required access agreed beforehand
- ✓No migration imposed by the troubleshooting
DirectAdmin
A problem on a DirectAdmin server?
On a DirectAdmin server, the incident rarely sits in the panel itself: it usually comes from the components it drives.
What we look at
- ✓CustomBuild
- ✓PHP and PHP-FPM
- ✓Apache and Nginx
- ✓MariaDB
- ✓DNS
- ✓SSL certificates
- ✓DirectAdmin services
- ✓Problems that appeared after an update
After the diagnosis
Sometimes repairing is not the best long-term answer
The diagnosis is what determines the right next step. In some cases a fix is enough; in others, the real answer lies elsewhere.
A server that has aged out
When fixes pile up on an end-of-life system, a migration sometimes makes more sense than one more repair.
Server migrationRecurring problems
When the same incidents keep coming back, ongoing support addresses the causes rather than the symptoms, one at a time.
Administration and managed servicesResources no longer sufficient
When the load durably exceeds what the server can deliver, no amount of tuning replaces properly sized resources.
Managed VPS Dedicated servers
Agencies and developers
Server support for web agencies and developers
We work on the server and the infrastructure. The application code stays with your team, and the diagnosis is precisely what establishes which side the problem is on.
Axanti also does web development: if the fix belongs to the application, we can talk about it, but that is never assumed up front.
Common situations
- ✓The developer knows the application, but not Linux administration
- ✓A deployment triggers a problem on the server side
- ✓The web server, PHP or the database has to be reviewed
- ✓The agency would rather not manage its client’s infrastructure
Security
Do you suspect a security incident?
Unusual behaviour, a suspected compromise, an exposed service or unexpected files deserve an analysis distinct from ordinary troubleshooting.
Axanti can review the environment, what the logs show and the state of the services, then determine the appropriate next step with you. We do not offer full forensic investigation, a guarantee that every piece of malicious code is removed, or a security certification: depending on what is found, rebuilding a clean environment is sometimes the safest route.
Before you write to us
What helps us start the diagnosis
The more precise the context, the faster the analysis starts. None of it is mandatory: describe what you know.
Never send a password or a private key through the form. The access required can be arranged appropriately after we are in touch.
Useful to mention
- ✓Type of server (VPS, dedicated, managed hosting)
- ✓Linux distribution, if you know it
- ✓Control panel, if any
- ✓Description of the problem
- ✓The exact error message, if you have one
- ✓Roughly when it started
- ✓Any recent change (update, deployment, configuration)
- ✓Hosting provider
FAQ
Frequently asked questions about server troubleshooting
The questions most often asked before handing us an incident to analyze.
Can you troubleshoot a server that is not hosted with Axanti?
Yes, when the necessary access is available and the environment allows it. The server can stay with your current provider: a troubleshooting job does not imply a migration.
Can I ask for a one-time intervention only?
Yes. No managed services contract is required to have a problem analyzed. The work covers the defined need, and recurring support remains possible afterwards if you want it.
Can you analyze a 500, 502, 503 or 504 error?
Yes. These codes signal a problem without naming its cause: it can come from the web server, PHP-FPM, a database, the available resources or the application itself. Reviewing the logs and the state of the services is what narrows it down.
Do you work on Nginx, Apache and PHP-FPM?
Yes: virtual hosts, reverse proxy, communication between the web server and PHP-FPM, PHP pools and versions, extensions, resource limits and errors that appeared after a version change.
Can you troubleshoot a DirectAdmin server?
Yes. DirectAdmin services, CustomBuild, PHP versions, the web stack, databases, DNS, certificates and email, including problems that appeared after an update or a configuration change.
Can you analyze a MariaDB or MySQL problem?
Yes: service unavailable, connection problems, load, configuration and errors that affect the applications. Recovering a corrupted database, however, depends on the actual state of the data and the available backups: it cannot be guaranteed.
What happens if the problem comes from the application rather than the server?
Telling those apart is exactly what the diagnosis is for. When the cause sits in the application code, we hand you what was observed on the server side so your developer can act on it. Axanti also does web development and can discuss that separately.
Do I have to send you my password in the form?
No. Never send a password, a private key or an access token through a form. Describe the problem and the context: the access needed will be arranged appropriately after we are in touch.
Is your server having a problem?
Describe the symptom, the services affected and what changed recently. We can then determine the best way to analyze the incident.