Server Administration
Day-to-day operation of Linux fleets: provisioning, patching, tuning and keeping the lights on.
- Provisioning & baseline config
- Patch & kernel management
- Apache / Nginx / PHP-FPM tuning
- Monitoring & alert design
Open to new opportunities
Alan Joy.
Software Engineer — Server Support
I keep production Linux infrastructure fast, hardened and online for a global managed-hosting client base. L2/L3 escalations, malware forensics and compromise recovery, InnoDB data recovery, and planned zero-downtime migrations on AWS — diagnosis through to an evidence-backed RCA.
0
Years in managed hosting
L2/L3
Escalation tier
0
Largest database migrated
45–80
Peak load average resolved
Alan Joy
Software Engineer — Server Support
About me
The engineer who gets the ticket after it's already been escalated twice.
I'm a software engineer working web hosting and server support at
Bobcares in Kochi, handling the managed-hosting side of
infrastructure — the place where a scanner burst without LVE caps, a
disk-full event or a rogue systemd service turns into a customer
outage within minutes.
Most of my day is L2 and L3 escalations across shared, VPS and
dedicated Linux servers for a global client base: reproducing the fault,
attributing load properly with pidstat and iostat instead of
guessing, applying a fix that holds, and writing the RCA the client actually
receives. The rest is planned work — malware forensics and compromise
recovery, database recovery, and migrations customers
shouldn't be able to feel.
I care about two things above all: not losing data, and not surprising the client. The PHP 5.6 → 8.2 migration I led ran header-based ALB routing so the new stack could be tested live before anyone was moved onto it, with a target-group rollback measured in seconds. I'm currently extending that into DevOps and cloud — AWS certifications in progress, GCP next.
RHEL / CentOS / Rocky
Ubuntu / Debian
cPanel / WHM
Plesk
DirectAdmin
CloudLinux & LVE
Apache / Nginx / LiteSpeed
MySQL / MariaDB
Exim / Postfix / Dovecot
ModSecurity / CSF
AWS EC2 / ALB / S3
Zabbix / Grafana
What I do
Services
Four things I do properly, rather than twenty things I do adequately.
L2 / L3 Hosting Support
Escalated tickets that L1 can't close — root-caused, fixed, and documented so they stay closed.
- Escalation handling & SLA response
- cPanel / WHM & Plesk deep issues
- Mail flow & deliverability repair
- Written root-cause analysis
Security & Incident Response
Hardening before the incident, and calm, methodical cleanup after one.
- CIS-aligned hardening baselines
- Malware & rootkit remediation
- Firewall, WAF & SSH lockdown
- Post-incident forensics report
Migrations & Upgrades
Moving live workloads between servers, panels or providers without customer-visible downtime.
- Server-to-server & panel migrations
- DNS cutover & TTL planning
- OS / PHP / MySQL version upgrades
- Tested rollback plan, every time
Toolbox
Technical skills
What I work with day to day across a global managed-hosting estate.
Control panels
cPanel / WHM
Plesk
DirectAdmin
CloudLinux & LVE
WHMCS
Softaculous
OS & web
RHEL
CentOS
Rocky Linux
Ubuntu
Debian
Apache
Nginx
LiteSpeed
PHP-FPM / CGI
LAMP & LEMP tuning
Databases
MySQL
MariaDB
InnoDB crash recovery
Corruption repair
Dump / restore
Slow-query tuning
Index tuning
Replication basics
Security
Malware forensics
Webshell / backdoor removal
ModSecurity (OWASP/Comodo)
CSF
iptables
firewalld
Fail2Ban
Imunify / maldet / ClamAV
KernelCare
SSH hardening
Security audits
Cloud & migration
AWS EC2
ALB
S3 / EBS
AWS Backup
SSM Session Manager
IAM
DigitalOcean
Cloudflare
Laravel Forge
GCP (learning)
Cutover & rollback planning
Mail & DNS
Exim
Postfix
Dovecot
SMTP / IMAP / POP3
SPF
DKIM
DMARC
Spam filtering
RBL delisting
SSL/TLS & AutoSSL
acme.sh
Monitoring & tools
Zabbix
Grafana
Nagios
sar / pidstat / iostat
top / atop
journalctl
Bash scripting
Git & Gitolite
rsync
PM2
Node.js services
Applications
WordPress
Joomla!
PrestaShop
Magento
WooCommerce
CS-Cart
DotNetNuke
Support workflow
WHMCS
Freshdesk
Tawk
SLA-driven ticketing
RCA documentation
Client reporting
Shift handover logs
Selected work
Incidents & projects
Real production problems I've owned end to end, from first alert to written RCA.
Migration
PHP 5.6 → 8.2 platform migration
Led the migration of a high-traffic AWS-hosted client to Ubuntu 24.04, Apache 2.4 and MariaDB 10.11 with a ~176 GB database — header-based ALB routing so the new stack could be tested live before any user was moved, full dump/restore sync, and a Gitolite auto-deploy pipeline.
→ Seconds-level target-group rollback path
Data recovery
WooCommerce recovery after InnoDB corruption
A disk-full event corrupted InnoDB and forced a MariaDB rebuild on a DirectAdmin server. Recovered order, tax, attribute and shipping data for multiple stores by working from pre-rebuild data-directory snapshots.
→ Live commerce data recovered, not written off
Incident
Rootkit-level compromise containment
Investigated and contained active compromises involving rogue systemd services, /etc/ld.so.preload hooks, hidden payload fetchers, webshells and crypto-mining processes — forensic log analysis, cleanup and post-incident hardening on WordPress and Joomla estates.
→ Contained, cleaned and hardened against repeat
From the blog
Notes from the terminal
Runbooks, post-mortems and hard-won lessons from managed hosting infrastructure.
Zero-Downtime cPanel Migration: A Field-Tested Runbook
The full sequence I follow to move live accounts between servers — pre-flight checks, TTL staging, delta syncs and the rollback plan you hope never to use.
Read runbookHardening a Fresh Linux Server: The First 30 Minutes
Everything I do between "the server exists" and "the server is safe to put traffic on" — SSH, firewall, kernel parameters, updates and audit logging.
Read checklistDiagnosing High Load Without Guessing
A repeatable triage order — load average, then CPU vs I/O vs memory, then the actual offending process — so you stop rebooting and start fixing.
Read the method
Next step
Got a server that needs looking after?
Whether it's an escalation you can't close, a migration you'd rather not do at 2am, or a fleet that's never been properly hardened — send me the details and I'll tell you honestly what I'd do.