Performance Tuning & Kernel Parameters (Sysctl)

Stop the kernel from throttling your network throughput

🧩 The Challenge

Dealing with a server that suddenly slows to a crawl when moving large backups or massive data sets is enough to make you pull your hair out. The default TCP window size settings haven’t been touched since the dial-up era and they are killing your modern high-speed backplane.

💡 The Fix

Crank up the TCP buffer sizes so your kernel can actually handle the throughput your network card is capable of. It’s an easy win that stops your NIC from sitting idle while waiting for packet acknowledgments.

sysctl -w net.core.rmem_max=16777216
sysctl -w net.core.wmem_max=16777216
sysctl -w net.ipv4.tcp_rmem='4096 87380 16777216'
sysctl -w net.ipv4.tcp_wmem='4096 65536 16777216'

⚙️ Why It Works

These parameters force the kernel to allocate much larger memory buffers for TCP sockets, allowing the sliding window to grow wide enough to keep data moving instead of waiting on the remote end to confirm receipt. You’re effectively telling the system to stop being so stingy with RAM when network pipes are wide open.

🚀 Pro-Tip: Don’t forget to append these to /etc/sysctl.conf unless you want to lose your sanity after the next reboot.

Linux Tips & Tricks | © ngelinux.com | 9/11/2026