Subject: Re: rt vs TODO? - msg#00136

Previous Message by Date: click to view message preview

Re: bytes > limit_maxbytes and no evictions?

So, instead of patching, answer me this. :) "Evictions" = if an item was removed and set with a TTL of > now(), right? If an item was set without a TTL, evictions is not incremented because we can calculate that? (total_items - curr_items - evictions) If so, then ignore my "patch" in the last email. :) - Steve -- Steve Webb - Lead System Administrator for Pronto.com Email: swebb@xxxxxxxxxx (Please send any work requests to: rt@xxxxxxxxxx) Cell: 303-564-4269, Office: 303-497-9367 YIM: scumola On Mon, 25 Jun 2007, Steve Webb wrote: Date: Mon, 25 Jun 2007 10:08:50 -0600 (MDT) From: Steve Webb <swebb@xxxxxxxxxx> To: memcached@xxxxxxxxxxxxxxx Subject: Re: bytes > limit_maxbytes and no evictions? Hey. So I checked out the code and "evictions" is only incremented when the data times out, which to me is not obvious. I changed the code to increment evictions when the cache is full and an item is "pushed" out. Just add the line: stats.evictions++; to items.c:437 after the stats.curr_items -= 1; line. Can someone with svn access please add this to the source tree? I don't have svn access and I need to be able to monitor this statistic and I'm guessing that others will also find this statistic interesting. We use it mainly to "tune" the size of a memcached server. A ton of evictions means that we need to increase the size of the server. Also, we chart this statistic, so a spike could signal a problem. - Steve

Next Message by Date: click to view message preview

Re: Problem installing 1.2.2 on Linux

it's odd that /usr/include isn't in your default include path. Are you mising 32-bit and 64-bit libraries perhaps? What version of Debian? What version of libevent? On Sat, Jun 23, 2007 at 06:27:51PM -0700, Nathan Schmidt wrote: > FWIW, this change lets memcached-1.2.2 compile and pass tests on > Debian -- my Makefile fu is not sufficient to properly remap the paths > that way but this is sufficient for my purposes. > > -Nathan > > +++ memcached-1.2.2-b/memcached.h 2007-06-24 01:11:24.774350680 +0000 > @@ -5,7 +5,10 @@ > #include <sys/types.h> > #include <sys/time.h> > #include <netinet/in.h> > -#include <event.h> > +#include </usr/include/event.h> > + > +typedef struct event_base; > +typedef struct event; > > #define DATA_BUFFER_SIZE 2048 > #define UDP_READ_BUFFER_SIZE 65536 > > > On 5/20/07, Don MacAskill <don@xxxxxxxxxxx> wrote: > > > > > >Ask Bjørn Hansen wrote: > >> > >> On May 20, 2007, at 1:33, Marcus Rueckert wrote: > >> > >>>> Red Hat Enterprise Linux ES release 3 (Taroon Update 8) > >>> > >>> redhat still supports that old versions? > >> > >> Yup - until 2010[1]. This is, in fact, one of the reasons RHEL is > >awesome. > >> > >> For the original question: I'm not sure if libevent uses epoll on RHEL3, > >> but I didn't have any problems compiling it and having it pass the tests. > >> > > > >Unless something has changed very recently, epoll isn't included by > >default in RHEL3, and it's a pain to patch in. I gave up. > > > >I believe it *is* included on RHEL4, or else it was very easy to > >install, since I don't remember our RHEL4 boxes requiring anything special. > > > >Don > > > -- Paul Lindner ||||| | | | | | | | | | lindner@xxxxxxxxx pgphKvyexJAMi.pgp Description: PGP signature

Previous Message by Thread: click to view message preview

rt vs TODO?

Where should feature requests for memcached go? I see 14 tickets in RT http://rt.livejournal.org/Search/Results.html?Query=Queue='Memcached%20*' (which are also listed in the top-level TODO.txt) and server/TODO which lists more items. Do we want to merge these or use one system? I think a link back to mailing list archives would be nice for some of these reports. -- Paul Lindner ||||| | | | | | | | | | lindner@xxxxxxxxx pgpLWKxT1H6Zf.pgp Description: PGP signature

Next Message by Thread: click to view message preview

Re: Largest production memcached install?

Have you considered using NAPI(Linux)/polling(FreeBSD)? Although it can increase response times a little, but helps freeing the machine from the interrupt load. On 05/03/07 20:48, Steve Grimm wrote: At peak times we see about 35-40% utilization (thatâs across all 4 CPUs.) But as you say, that number will vary dramatically depending on how you use it. The biggest single user of CPU time isnât actually memcached per se; itâs interrupt handling for all the incoming packets. -Steve On 5/3/07 11:41 AM, "Jerry Maldonado" <jerry.maldonado@xxxxxxxxxxxxx> wrote: With the configuration you noted below, what is your CPU utilization. We are implementing memcached in our environment and I am trying to get a feel for what we will need for production. I realize that it all depends on how we are using it, but I am interested to see what it is based on your configuration. Thanks, Jerry -----Original Message----- *From:* memcached-bounces@xxxxxxxxxxxxxxx [mailto:memcached-bounces@xxxxxxxxxxxxxxx] <mailto:memcached-bounces@xxxxxxxxxxxxxxx%5D>*On Behalf Of *Steve Grimm *Sent:* Thursday, May 03, 2007 11:33 AM *To:* Sam Lavery; memcached@xxxxxxxxxxxxxxx *Subject:* Re: Largest production memcached install? No clue if weâre the largest installation, but Facebook has roughly 200 dedicated memcached servers in its production environment, plus a small number of others for development and so on. A few of those 200 are hot spares. They are all 16GB 4-core AMD64 boxes, just because thatâs where the price/performance sweet spot is for us right now (though it looks like 32GB boxes are getting more economical lately, so I suspect weâll roll out some of those this year.) We have a home-built management and monitoring system that keeps track of all our servers, both memcached and other custom backend stuff. Some of our other backend services are written memcached-style with fully interchangeable instances; for such services, the monitoring system knows how to take a hot spare and swap it into place when a live server has a failure. When one of our memcached servers dies, a replacement is always up and running in under a minute. All our services use a unified database-backed configuration scheme which has a Web front-end we use for manual operations like adding servers to handle increased load. Unfortunately that management and configuration system is highly tailored to our particular environment, but I expect you could accomplish something similar on the monitoring side using Nagios or another such app. All that said, I agree with the earlier comment on this list: start small to get some experience running memcached in a production environment. Itâs easy enough to expand later once you have appropriate expertise and code in place to make things run smoothly. -Steve On 5/3/07 8:06 AM, "Sam Lavery" <sam.lavery@xxxxxxxxx> wrote: Does anyone know what the largest installation of memcached currently is? I'm considering putting it on 100+ machines(solaris/mod_perl), and would love to hear any tips people have for managing a group of that size(and larger). Additionally, are there any particular patches I should try out for this specific platform? Thanks in advance, Sam -- Attila Nagy e-mail: Attila.Nagy@xxxxxx Free Software Network (FSN.HU) phone: +3630 306 6758 http://www.fsn.hu/