Bug 32041
| Summary: | Add point about net.inet.tcp.portange.{first,last} sysctl tunable to tuning(7) manpage | ||||||
|---|---|---|---|---|---|---|---|
| Product: | Documentation | Reporter: | mark | ||||
| Component: | Books & Articles | Assignee: | freebsd-doc (Nobody) <doc> | ||||
| Status: | Closed FIXED | ||||||
| Severity: | Affects Only Me | ||||||
| Priority: | Normal | ||||||
| Version: | Latest | ||||||
| Hardware: | Any | ||||||
| OS: | Any | ||||||
| Attachments: |
|
||||||
State Changed From-To: open->analyzed Good point, that probably was the reason the benchmark failed at 3000 connections. I'll commit this patch (with a few additions) soon. State Changed From-To: analyzed->closed Closed by request, recent changes to tuning(7) covers the port range issue nicely. |
tuning(7) doesn't mention the net.inet.tcp.portange.{first,last} sysctl tunables and probably ought to as at least one published benchmark failed because the benchmarkers were unaware of its existence. http://www.samag.com/documents/s=1147/sam0108q/ How-To-Repeat: Read the man page and fail to realize that net.inet.tcp.portange.{first,last} might need adjustment.