About

I am Attila Csontos. I work on infrastructure: Linux, Docker Swarm, monitoring, on-premise deployments and the long tail of tooling that surrounds them.

This site exists because of a specific frustration. When a system throws Input/output error (os error 5) or subsystem request failed on channel 0, a search usually returns three kinds of pages: someone with the same problem and no answer, a generic tutorial that never mentions the error, or a page that lists five “solutions” that the author clearly never ran.

So the rule here is narrow:

• Every post starts from an error message or symptom I hit myself, quoted exactly as the system printed it.
• The diagnosis section shows what I measured, not what I assumed — command output, headers, exit codes.
• The fix is the one that actually resolved it on the affected machine, and I say so explicitly when something is a workaround rather than a fix.
• When I do not know the root cause, the post says that instead of inventing one.

If a post is wrong or incomplete on your setup, tell me — see Contact. Corrections get folded into the post with a note about what changed.