ROOT@PRODUCTION  ::  MIDDLEWARE / UNIX

I keep middleware alive when everything is trying to break it.Petar — Linux / Unix Systems Administrator, WebSphere & Apache middleware, moving into SRE

I operate at the layer most engineers never see: WebSphere, IHS, Apache, and Tomcat under real production load, across multi-region AWS infrastructure. My work is diagnosing the failures other people give up on — race conditions, SELinux context mismatches, false-positive security scans — and turning them into documented, repeatable fixes.

> VIEW_LOG> OPEN_UPLINK

// KEY_METRICS

STATUS: 200 OK
1
full DR failover validated end-to-end on AWS EMEA — bug found and fixed before it could hit a real incident
0
recurring IHS duplicate-startup failures, down from a persistent boot-time race condition
RHEL_8/9
Apache hardening baseline shipped — TLS, SELinux, firewalld, systemd, security headers
1
false-positive security scan disputed and overturned, protecting client compliance standing
SELINUX
policy conflicts resolved by fixing contexts, never by disabling enforcement
4
middleware platforms operated in production: WebSphere, IHS, Apache, Tomcat

// INCIDENT_LOG

5 ENTRIES
~/aws-emea/dr-validation
01AWS EMEA Disaster Recovery Validation
problem
DR environment for a production WebSphere system had never been failover-tested. Unknown whether it would survive a real incident.
solution
Ran a live DR test, traced a BindException to a hostname/IP mismatch on the DR EC2 hosts, corrected via /etc/hosts, and wrote a structured, repeatable DR runbook.
impact
Failure found and fixed in a drill, not in an outage. DR posture went from untested to validated and documented.
~/ihs/start_ihs_safe.sh
02IHS Startup Race Condition
problem
IHS intermittently starting twice on boot — crontab race condition from chained && reboot entries — causing port conflicts.
solution
Diagnosed the sequencing bug and wrote start_ihs_safe.sh, a startup guard preventing duplicate initialization.
impact
Eliminated a recurring, unpredictable boot-time failure class without touching platform orchestration.
~/rhel/apache-hardening
03RHEL 8/9 Apache Hardening Baseline
problem
No standardized hardening baseline for Apache across the fleet — inconsistent exposure to CVEs and audit findings.
solution
Built a hardening guide: module minimization, TLS config, security headers, SELinux policy, firewalld rules, systemd drop-in overrides.
impact
Reusable, auditable hardening standard for every RHEL 8/9 Apache instance — a direct foundation for CIS-aligned compliance.
~/selinux/control-m-policy
04SELinux Policy Conflict — Control-M
problem
Control-M agent service failing silently — SELinux context mismatch, bin_t vs. shell_exec_t.
solution
Traced the mismatch through SELinux audit logs and corrected the file context policy.
impact
Restored critical job-scheduling integration with enforcement fully on — fixed, not bypassed.
~/tls/scan-dispute
05Security Scan Dispute & Vendor Validation
problem
Third-party testssl.sh scan reported findings caused by a bash socket-fallback artifact, not a real vulnerability.
solution
Investigated the scanner's execution mode, identified the false-positive mechanism, drafted the technical rebuttal.
impact
Avoided an unnecessary remediation cycle and protected the team's standing on a compliance-sensitive finding.

// CORE_STACK

SYSADMIN → SRE
IaC / AnsibleLEARNING
Manual config management across WAS, IHS, Tomcat — consistent enough to codify.
Converting proven runbooks into Ansible roles
CI/CDLEARNING
Diagnosed a Jenkins SSH auth failure on a live deployment pipeline.
GitHub Actions / GitLab CI for middleware deploys
ObservabilityLEARNING
Log-driven root-causing — SELinux audit, Jenkins, DR bind errors.
Prometheus + Grafana for proactive detection
Security & HardeningACTIVE
CIS-adjacent Apache/SELinux/firewalld hardening, TLS negotiation debugging.
Scored CIS benchmarks, automated compliance checks