Browse categories
English
$
USD
Connect to your AI
I will fix your self hosted ai, docker or linux server
About this Gig
Something is broken and the forum threads are not helping. I find the root cause and fix it, live, while you watch.
Ten years of enterprise support, most recently as designated engineer for Fortune 500 and government accounts. I work from evidence: logs, config, reproduction, then the fix. No reinstalling and hoping.
What I fix
- Linux and UNIX: Ubuntu, Debian, RHEL, Rocky, Solaris, AIX, HP-UX. Boot, panics, file-system corruption, memory, performance, services, SSH, TLS
- Windows Server: services, Active Directory, DNS, SQL Server, Exchange, Hyper-V
- Virtualisation and storage: VMware, Proxmox, SAN, iSCSI, NAS, tape
- Docker, Compose, Kubernetes: crashes, restart loops, networking, volumes, pods pending
- Self-hosted AI: Ollama, llama.cpp, vLLM, LiteLLM, n8n, vector DBs, GPU passthrough, CUDA
- Backups that fail or will not restore: NetBackup, Backup Exec, Enterprise Vault
- Networks: pfSense, firewalls, VLANs, DNS
How it works: send symptom and logs when you order. Basic: 30-min call, root cause in writing plus a fix plan. Standard and Premium: 60 or 90-min sessions over screen-share or SSH where I fix it, with a write-up.
Bigger jobs: custom offer. Urgent? Message first
FAQ
What access do you need?
For the Basic call, logs and config you attach are usually enough. For a fix session, a screen-share or temporary SSH access that you watch and revoke afterwards. I never need your root password stored anywhere.
Which operating systems and platforms?
Linux (Ubuntu, Debian, RHEL, Rocky), Solaris, AIX, HP-UX, Windows Server with Active Directory, DNS, SQL Server and Exchange, VMware, Hyper-V, Proxmox, Docker, Kubernetes, AWS, Azure and GCP, plus storage: SAN, iSCSI, NAS and tape.
What if the root cause is hardware or a bug you cannot fix?
You still get the written root cause and your options, which is what the Basic call covers. If a fix is out of reach, I say so before you spend more.
How fast can you start?
Usually the same day. If it is production-down, message me first and add the fast-delivery extra when you order.
What if it takes longer than the session?
Most single issues fit in 60 minutes once the logs are in hand. If it does not, you get the written root cause and a fix plan at the end of the session, and a follow-up session is available as an extra.
