# Why SSH Says client\_loop send disconnect Broken pipe *And How to Kill It...With Extreme Prejudice* ### TL;DR A “pipe,” in this context, is a metaphor for any data channel most often a network socket for SSH. When you see **client\_loop: send disconnect: Broken pipe**, SSH just tried to send your precious bits to a connection long gone. Fix it with proper keepalive settings. And remember: lost connections shouldn’t tank your automation wrap, retry, and keep calm. *** **“Broken pipe.”** The error haunts SSH users everywhere. You’re in the middle of a remote job, deep into tweaks or long-running automation, and BOOM session gone. Time to pull back the curtain: ### What Is a “Pipe” (and Why Is It Broken)? * **Classic Unix Pipe:** * A pipe is an in-memory channel one process writes, one reads. Removed in real-time, never hits the disk. * **Network Sockets (What SSH Actually Uses):** * *Pipes made network-aware*: A socket is a two-way stream connecting apps, often over TCP, letting data flow between devices. * **Under the Hood:** * When a program tries to write to a pipe (or socket) and the other end has bailed (closed, crashed, lost the network), Unix gives the process a "broken pipe" (`SIGPIPE`). The next write dead on arrival[^1]. > "**Broken pipe** here is Unix’s way of telling you: ‘I yelled, but nobody’s listening on the other side anymore.’ Not a polite goodbye more like a door slammed shut.” ### The Real Reason SSH Drops the Connection #### Why the “Broken Pipe” Triggers in SSH: * **Inactivity Timeout:** No commands, no network traffic SSH server closes idle connections to save resources. * **Network Hiccups:** Packet loss, WiFi flickers, VPN drops, or bad routing? Session gets chopped. * **Client/Server Keepalive Misconfig:** If either side expects regular “I’m alive!” pings (but doesn’t get them), it’s curtains for your session[^2]. * **Resource Problems & Network Firewall Timeouts**: ISPs or corporate firewalls can kill dormant connections faster than you can say `Ctrl+C`. ### "Broken Pipe" = Socket Graveyard * **SSH Is Built On Sockets:** The error lingo is legacy; modern SSH is about sockets, not shell pipes, but the OS’s behavior and error messages are the same[^1]. * **Result:** Your SSH client tries to send data but gets the cold shoulder connection is toast. ### Real-World Ripples * **Lone SSH Client Dies:** The connection is gone, but if you use `tmux` or `screen`, your remote stuff keeps running. Reconnect, reattach, no harm done. * **Automation Alert:** Pushing files or running scripts on remote servers? *Broken pipe* can cause partial uploads, failed steps, and very messy ops. ### How to NEVER See "Broken Pipe" Again 🚀 **The fast fix:** Make SSH keep the connection “breathing,” even when you’re AFK. #### Client-Side Fix Edit (or create) your `~/.ssh/config`: ``` Host * ServerAliveInterval 60 ServerAliveCountMax 5 ``` Or add the options to your SSH call: ``` ssh -o ServerAliveInterval=60 -o ServerAliveCountMax=5 user@host ``` * `ServerAliveInterval`: (seconds) How often the client pings the server[^3]. * `ServerAliveCountMax`: How many pings the client will try before giving up. * **Example:** 5 tries × 60sec = 5 minutes of trying before disconnect[^4]. #### Server-Side Fix With root, in `/etc/ssh/sshd_config`: ``` ClientAliveInterval 60 ClientAliveCountMax 5 ``` Then: ``` sudo systemctl restart sshd ``` * `ClientAliveInterval` & `ClientAliveCountMax`: Similar concept, but from the server’s point of view how long it’ll wait and how many times it’ll shout into the void before killing the session[^5]. ### For the Hackers, Ops and Defence * **Automation-Ready:** Always wrap file transfers or infra ops in retry logic, so a “broken pipe” doesn’t break your CI/CD pipeline. * **Debug Like a Pro:** * Use `ssh -vvv` for deep diagnostics. * Count how long until the disconnect, pinpoint which timeout is killing you. * **Cloud/NAT Environments:** Some cloud networks drop *silent* TCP connections; tune your keepalives to run just under the timeout trigger. ⁂ [^1]: [^2]: [^3]: [^4]: [^5]: --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.wehost.co.in/blog/why-ssh-says-client_loop-send-disconnect-broken-pipe.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.