------------------------------------------------------------ revno: 472 revision-id: [EMAIL PROTECTED] parent: [EMAIL PROTECTED] parent: [EMAIL PROTECTED] committer: Andrew Tridgell <[EMAIL PROTECTED]> branch nick: tridge timestamp: Wed 2007-06-06 10:29:27 +1000 message: merged vsftpd event script from ronnie added: config/events.d/40.vsftpd 40.vsftpd-20070605080716-mih15mjbgsjt7niv-1 modified: config/events.d/59.nfslock nfslock-20070601105340-vlcvnp6euoj3zdwy-2 packaging/RPM/ctdb.spec ctdb.spec-20070527204758-biuh7znabuwan3zn-3 tools/ctdb_control.c ctdb_control.c-20070426122705-9ehj1l5lu2gn9kuj-1 web/ctdb.html ctdb.html-20070601052353-vgod9lfo4an4o83j-2 ------------------------------------------------------------ revno: 432.1.27 merged: [EMAIL PROTECTED] parent: [EMAIL PROTECTED] committer: Ronnie Sahlberg <[EMAIL PROTECTED]> branch nick: ctdb timestamp: Tue 2007-06-05 18:42:49 +1000 message: ctdb is only a ha solution when combined with a cluster filesystem ------------------------------------------------------------ revno: 432.1.26 merged: [EMAIL PROTECTED] parent: [EMAIL PROTECTED] committer: Ronnie Sahlberg <[EMAIL PROTECTED]> branch nick: ctdb timestamp: Tue 2007-06-05 18:40:43 +1000 message: add some text about CTDB and in which scenarios it would be a good candidate to solve the users needs for a HA solution ------------------------------------------------------------ revno: 432.1.25 merged: [EMAIL PROTECTED] parent: [EMAIL PROTECTED] committer: Ronnie Sahlberg <[EMAIL PROTECTED]> branch nick: ctdb timestamp: Tue 2007-06-05 18:32:06 +1000 message: provide machinereadable output for ctdb ip ------------------------------------------------------------ revno: 432.1.24 merged: [EMAIL PROTECTED] parent: [EMAIL PROTECTED] parent: [EMAIL PROTECTED] committer: Ronnie Sahlberg <[EMAIL PROTECTED]> branch nick: ctdb timestamp: Tue 2007-06-05 18:16:45 +1000 message: merge from tridge ------------------------------------------------------------ revno: 432.1.23 merged: [EMAIL PROTECTED] parent: [EMAIL PROTECTED] committer: Ronnie Sahlberg <[EMAIL PROTECTED]> branch nick: ctdb timestamp: Tue 2007-06-05 18:14:01 +1000 message: add a simple events script to manage vsftpd ------------------------------------------------------------ revno: 432.1.22 merged: [EMAIL PROTECTED] parent: [EMAIL PROTECTED] committer: Ronnie Sahlberg <[EMAIL PROTECTED]> branch nick: ctdb timestamp: Tue 2007-06-05 12:43:35 +1000 message: dont use CTDB_MANAGES_NFS for controlling the lockmanager use a dedicated variable CTDB_MANAGES_NFSLOCK since some might want to use nfs but no lockmanager ------------------------------------------------------------ revno: 432.1.21 merged: [EMAIL PROTECTED] parent: [EMAIL PROTECTED] parent: [EMAIL PROTECTED] committer: Ronnie Sahlberg <[EMAIL PROTECTED]> branch nick: ctdb timestamp: Tue 2007-06-05 08:43:30 +1000 message: merge from tridge === added file 'config/events.d/40.vsftpd' --- a/config/events.d/40.vsftpd 1970-01-01 00:00:00 +0000 +++ b/config/events.d/40.vsftpd 2007-06-05 08:14:01 +0000 @@ -0,0 +1,45 @@ +#!/bin/sh +# event strict to manage vsftpd in a cluster environment + +. /etc/ctdb/functions +loadconfig vsftpd + +[ "$CTDB_MANAGES_VSFTPD" = "yes" ] || exit 0 + +cmd="$1" +shift + +case $cmd in + startup) + /bin/mkdir -p /etc/ctdb/state/vsftpd + + # make sure the service is stopped first + service vsftpd stop > /dev/null 2>&1 + service vsftpd start + ;; + + shutdown) + service vsftpd stop + ;; + + takeip) + echo "restart" >> /etc/ctdb/state/vsftpd/restart + ;; + + releaseip) + echo "restart" >> /etc/ctdb/state/vsftpd/restart + ;; + + recovered) + # if we have taken or released any ips we must + # restart vsftpd to ensure that all tcp connections are reset + [ -f /etc/ctdb/state/vsftpd/restart ] && { + service vsftpd stop > /dev/null 2>&1 + service vsftpd start + } >/dev/null 2>&1 + + /bin/rm -f /etc/ctdb/state/vsftpd/restart + ;; +esac + +exit 0
=== modified file 'config/events.d/59.nfslock' --- a/config/events.d/59.nfslock 2007-06-05 05:18:37 +0000 +++ b/config/events.d/59.nfslock 2007-06-05 08:16:45 +0000 @@ -4,7 +4,7 @@ . /etc/ctdb/functions loadconfig nfs -[ "$CTDB_MANAGES_NFS" = "yes" ] || exit 0 +[ "$CTDB_MANAGES_NFSLOCK" = "yes" ] || exit 0 [ -z "$STATD_SHARED_DIRECTORY" ] && exit 0 === modified file 'packaging/RPM/ctdb.spec' --- a/packaging/RPM/ctdb.spec 2007-06-04 05:09:03 +0000 +++ b/packaging/RPM/ctdb.spec 2007-06-05 08:14:01 +0000 @@ -94,6 +94,7 @@ %{_sysconfdir}/ctdb/events %{_sysconfdir}/ctdb/functions %{_sysconfdir}/ctdb/events.d/10.interface +%{_sysconfdir}/ctdb/events.d/40.vsftpd %{_sysconfdir}/ctdb/events.d/50.samba %{_sysconfdir}/ctdb/events.d/59.nfslock %{_sysconfdir}/ctdb/events.d/60.nfs === modified file 'tools/ctdb_control.c' --- a/tools/ctdb_control.c 2007-06-06 00:25:46 +0000 +++ b/tools/ctdb_control.c 2007-06-06 00:29:27 +0000 @@ -355,6 +355,17 @@ return ret; } + if(options.machinereadable){ + printf(":Public IP:Node:\n"); + for(i=0;i<ips->num;i++){ + printf(":%s:%d:\n", + inet_ntoa(ips->ips[i].sin.sin_addr), + ips->ips[i].takeover_vnn); + } + return 0; + } + + printf("Number of nodes:%d\n", ips->num); for(i=0;i<ips->num;i++){ printf("%-16s %d\n", === modified file 'web/ctdb.html' --- a/web/ctdb.html 2007-06-03 06:33:08 +0000 +++ b/web/ctdb.html 2007-06-05 08:42:49 +0000 @@ -81,6 +81,19 @@ seconds. During the recovery period the databases are 'frozen', and all IO operations by ctdb clients are suspended. +<h3>Is CTDB a HA solution?</h3> +Yes and no.<br> +CTDB alone is not a HA solution, but when you combine CTDB with a clustered +filesystem it becomes one.<br> +CTDB is primarily developed around the concept of having a shared cluster +filesystem shared across all the nodes in the cluster to provide the features +required for building a NAS cluster.<br> +Thus CTDB relies on an external component (the cluster filesystem) to provide +the mechanisms for avoiding split-brain and other difficult tasks.<br><br> +However, If you applications do need a clustered filesystem for all the nodes, +in that scenario CTDB will provide a very easy to install and manage +solution for your clustering HA needs. + <h3>IP Takeover</h3> When a node in a cluster fails, CTDB will arrange that a different