| Rev. | 时间 | 作者 | Message |
|---|---|---|---|
| r14 | 2010-10-01 20:33:45 | tullio | featuring delete command (under testing) |
| r13 | 2010-07-27 15:15:35 | tullio | Preserved expire time |
| r12 | 2010-07-22 21:24:45 | tullio | changed SIGHUP handling to SIGINT handling |
| r11 | 2010-07-20 20:21:42 | tullio | removed debug codes |
| r10 | 2010-07-20 20:21:25 | tullio | fixed remained bug calculating data length |
| r9 | 2010-07-20 19:54:40 | tullio | fixed measuring data length bug |
| r8 | 2010-07-20 15:14:27 | tullio | checked if under replication mode or not |
| r7 | 2010-07-19 21:19:36 | tullio | modified Makefile.in to output final executable from "mem... |
| r6 | 2010-07-16 20:21:53 | tullio | now destination hosts should be written in a file |
| r5 | 2010-07-15 00:24:20 | tullio | pool version and added benchmark program for FreeBSD |
ForkDependencies:
-- libevent, http://www.monkey.org/~provos/libevent/ (libevent-dev)
If using Linux, you need a kernel with epoll. Sure, libevent will
work with normal select, but it sucks.
epoll isn't in Linux 2.4, but there's a backport at:
http://www.xmailserver.org/linux-patches/nio-improve.html
You want the epoll-lt patch (level-triggered).
If you're using MacOS, you'll want libevent 1.1 or higher to deal with
a kqueue bug.
Also, be warned that the -k (mlockall) option to memcached might be
dangerous when using a large cache. Just make sure the memcached machines
don't swap. memcached does non-blocking network I/O, but not disk. (it
should never go to disk, or you've lost the whole point of it)
The memcached website is at:
http://www.memcached.org
Want to contribute? Up-to-date pointers should be at:
http://contributing.appspot.com/memcached