Jul 29, 2009
Wie kann ich mehrere Instanzen von memcached auf einem Server laufen lassen?
Problem
Ich brauche mehr als einen Memcached Server.
Lösung
Virtuellemaschinen
Ich kann für jeden memcached server eine Virtuellemaschine erstellen.
Probleme
- mehr speicher (system+memcached)
- eine ip addresse pro Instanz
- mehr Aufwand
Multiple Instanzen mithilfe von Ports
Dafür müssen die Skripte init.d und start-memcached gepatched werden um Multiple Instanzen zum laufen zubekommen.
Nach dem patch sucht es nach Konfigurationsdateien die folgendem Muster entsprechen /etc/memcached_*.conf. Wenn es keine Dateien findet benutzt es die /etc/memcached.conf Konfigurationsdatei. Es ist also rückwärtskompatibel.
Wenn du nun 2 memached Instanzen erstellen möchtest kopiere die Konfigurationsdatei und ändere den Port, Speicher …
cp /etc/memcached.conf /etc/memcached_myserver_1.conf cp /etc/memcached.conf /etc/memcached_myserver_2.conf
Befehle
startet alle Server
/etc/init.d/memcached start
startet nur einen bestimmten Server
/etc/init.d/memcached start myserver_1
stopt alle Server
/etc/init.d/memcached stop
stopt nur einen bestimmten Server
/etc/init.d/memcached stop myserver_1
Die Änderungen
/usr/share/memcached/scripts/start-memcached
26,30d25 > if (scalar(@ARGV) == 2) { > $etcfile = shift(@ARGV); > $pidfile = shift(@ARGV); > } >
/etc/init.d/memcached
16a17 > DAEMONNAME=memcached 18d18 < NAME=memcached 20d19 < PIDFILE=/var/run/$NAME.pid 26a26,63 > FILES=(/etc/memcached_*.conf); > # check for alternative config schema > if [ -r "${FILES[0]}" ]; then > CONFIGS=(); > for FILE in "${FILES[@]}"; > do > # remove prefix > NAME=${FILE#/etc/}; > # remove suffix > NAME=${NAME%.conf}; > > # check optional second param > if [ $# -ne 2 ]; > then > # add to config array > CONFIGS+=($NAME); > elif [ "memcached_$2" == "$NAME" ]; > then > # use only one memcached > CONFIGS=($NAME); > break; > fi; > done; > > if [ ${#CONFIGS[@]} == 0 ]; > then > echo "Config not exist for: $2" >&2; > exit 1; > fi; > else > CONFIGS=(memcached); > fi; > > CONFIG_NUM=${#CONFIGS[@]}; > for ((i=0; i < $CONFIG_NUM; i++)); do > NAME=${CONFIGS[${i}]}; > PIDFILE="/var/run/${NAME}.pid"; > 30c67 < start-stop-daemon --start --quiet --exec $DAEMONBOOTSTRAP --- > start-stop-daemon --start --quiet --exec "$DAEMONBOOTSTRAP" -- /etc/${NAME}.conf $PIDFILE 50c87 < start-stop-daemon --start --quiet --exec $DAEMONBOOTSTRAP --- > start-stop-daemon --start --quiet --exec "$DAEMONBOOTSTRAP" -- /etc/${NAME}.conf $PIDFILE 54c91 < N=/etc/init.d/$NAME --- > N=/etc/init.d/$DAEMONNAME 59a97 > done;
No related posts.
Ähnliche Artikel bereitgestellt von Yet Another Related Posts Plugin.











Could you post the full files or email me the 2 files? I’m having some problems when trying to run the patch
FYI
Make sure you check line 1 on /etc/init.d/memcached
it should be :
#! /bin/bash
not
#! /bin/sh
else the code will not work.
Very nice,
thanks
Welche Version von Memcached wurde hier gepatched? Ich arbeite gerade an 1.4.2. und es scheint nicht zusammen zu passen.
Unfortunately i can’t get that to work. How do you apply the patch files? I tried it by hand but there seems to be an error. all i get is a
/etc/init.d/memcached: line 100: syntax error near unexpected token `;’
that’s weird. I should have patched this automatically but this isn’t a diff-file is it?
Looks like the patch has a tiny syntax bug-
> CONFIG_NUM=${#CONFIGS[@]};
> for ((i=0; i NAME=${CONFIGS[${i}]};
> PIDFILE=”/var/run/${NAME}.pid”;
should be
> CONFIG_NUM=${#CONFIGS[@]};
> for ((i=0; i NAME=${CONFIGS[${i}]};
> PIDFILE=”/var/run/${NAME}.pid”;
> done
Note the done
have you apply this patch? i think the “done” is in line 59
Hello,
How is one supposed to use the code provided in the patch windows?
Regards
Andrew
You can use the command patch or manually change the files.
I had googled this page: http://linux.about.com/od/commands/l/blcmdl1_patch.htm
I am new to using the patch command so bear with me! I created a file on my server called diff1 containing the code for /usr/share/memcached/scripts/start-memcached. I then created a backup of my start-memcached. When I run patch start-memcached.bak < diff1, I receive this error message:
patching file start-memcached.bak
patch: **** `<' expected at line 2 of patch
Am I doing something stupid?
Regards
Andrew
which version of memcache you use?
I am currently using memcached 1.4.2-1ubuntu3 and these patches will not work for me at all. on the first patch I keep getting this error: patch: **** ` was expected, this was caused by there not being a space after those such as ‘> ‘. I fixed that and got the above errors.
Any assistance?
Joseph Crawford´s last [type] ..Repairing the Oil Pan on my 2006 Katana 600
#full /etc/init.d/memcached for my ubuntu 10.04+
#! /bin/bash
### BEGIN INIT INFO
# Provides: memcached
# Required-Start: $syslog
# Required-Stop: $syslog
# Should-Start: $local_fs
# Should-Stop: $local_fs
# Default-Start: 2 3 4 5
# Default-Stop: 0 1 6
# Short-Description: memcached – Memory caching daemon
# Description: memcached – Memory caching daemon
### END INIT INFO
PATH=/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin
DAEMON=/usr/bin/memcached
DAEMONBOOTSTRAP=/usr/share/memcached/scripts/start-memcached
DAEMONNAME=memcached
DESC=memcached
PIDFILE=/var/run/$NAME.pid
FILES=(/etc/memcached_*.conf);
if [ -r "${FILES[0]}” ]; then
CONFIGS=();
for FILE in “${FILES[@]}”;
do
# remove prefix
NAME=${FILE#/etc/};
# remove suffix
NAME=${NAME%.conf};
# check optional second param
if [ $# -ne 2 ];
then
# add to config array
CONFIGS+=($NAME);
elif [ "memcached_$2" == "$NAME" ];
then
# use only one memcached
CONFIGS=($NAME);
break;
fi;
done;
if [ ${#CONFIGS[@]} == 0 ];
then
echo “Config not exist for: $2″ >&2;
exit 1;
fi;
else
CONFIGS=(memcached);
fi;
CONFIG_NUM=${#CONFIGS[@]};
for ((i=0; i &2
exit 1
;;
esac
exit 0
Here are the complete patched files for Ubuntu 8.04.
http://gist.github.com/516746
Awesome tip, thank very much for the info!
small Fix:
use:
FILES=(/etc/memcached*.conf);
instead of:
FILES=(/etc/memcached_*.conf);
to also hit the standard-configfile
and simply add the line:
start-stop-daemon –stop –quiet –oknodo –pidfile $PIDFILE
whenever a serverstop is invoked.
forget the “start-stop-daemon –stop –quiet –oknodo –pidfile $PIDFILE” Line cause its already there
i have forked the gist and make some small improvements http://gist.github.com/612961.
tested under:
ubuntu and memcache 1.4.5
debian lenny memcache 1.4.5
You guys are lifesavers, I didn’t want to have to write this myself!
Works perfectly!!!!!!
Thanks for your scripts, I have tested it in Debian lenny – memcached 1.4.5-1 and Works perfectly.
But, It would be possible to run more than 2 instances of memcache?.
If so, what should I change in the scripts?
simple add another memcache config file. you are only limited on ram and ports
so i run 22 memcache instances on one server
Thanks for your answer!
In this regard. Can I do some questions?
1. Do you know any tool that allows me to make monitoring multiple instances at once?
2. Do you know any tool that allows me to make “Stress Testing” on a site configured with memcache, which only has access through a “logon”?
Thanks in advance!!
oh cool, this is great! just like what i needed! thanks so much!
april14344´s last [type] ..Young Driver Insurance – The Best Strategy to Save You Money!