1
0
mirror of https://github.com/samba-team/samba.git synced 2024-12-23 17:34:34 +03:00

remove the obsolete ipmux component.

this is replaced by LVS since a long time

(This used to be ctdb commit dca41ec04788922ce5f4c52d346872b3e35f8cbb)
This commit is contained in:
Ronnie Sahlberg 2009-05-25 12:33:52 +10:00
parent 7b163bca18
commit caf0e863a4
10 changed files with 35 additions and 522 deletions

View File

@ -26,8 +26,6 @@ POPT_LIBS = @POPT_LIBS@
POPT_CFLAGS = @POPT_CFLAGS@
POPT_OBJ = @POPT_OBJ@
IPQ_LIBS = @IPQ_LIBS@
CFLAGS=-g -I$(srcdir)/include -Iinclude -Ilib -Ilib/util -I$(srcdir) \
-I@tallocdir@ -I@tdbdir@/include -I@libreplacedir@ \
-DVARDIR=\"$(localstatedir)\" -DETCDIR=\"$(etcdir)\" \
@ -63,7 +61,7 @@ TEST_BINS=tests/bin/ctdb_bench tests/bin/ctdb_fetch tests/bin/ctdb_store \
tests/bin/ctdb_traverse tests/bin/rb_test tests/bin/ctdb_transaction \
@INFINIBAND_BINS@
BINS = bin/ctdb @CTDB_SCSI_IO@ bin/ctdb_ipmux bin/smnotify bin/ping_pong
BINS = bin/ctdb @CTDB_SCSI_IO@ bin/smnotify bin/ping_pong
SBINS = bin/ctdbd
DIRS = lib bin tests/bin
@ -104,10 +102,6 @@ bin/scsi_io: $(CTDB_CLIENT_OBJ) utils/scsi_io/scsi_io.o
@echo Linking $@
@$(CC) $(CFLAGS) -o $@ utils/scsi_io/scsi_io.o $(CTDB_CLIENT_OBJ) $(LIB_FLAGS)
bin/ctdb_ipmux: $(CTDB_CLIENT_OBJ) utils/ipmux/ipmux.o
@echo Linking $@
@$(CC) $(CFLAGS) -o $@ utils/ipmux/ipmux.o $(CTDB_CLIENT_OBJ) $(LIB_FLAGS) $(IPQ_LIBS)
bin/ctdb: $(CTDB_CLIENT_OBJ) tools/ctdb.o tools/ctdb_vacuum.o
@echo Linking $@
@$(CC) $(CFLAGS) -o $@ tools/ctdb.o tools/ctdb_vacuum.o $(CTDB_CLIENT_OBJ) $(LIB_FLAGS)
@ -202,7 +196,6 @@ install: all
mkdir -p $(DESTDIR)$(docdir)/ctdb
${INSTALLCMD} -m 644 ctdb.pc $(DESTDIR)$(libdir)/pkgconfig
${INSTALLCMD} -m 755 bin/ctdb $(DESTDIR)$(bindir)
${INSTALLCMD} -m 755 bin/ctdb_ipmux $(DESTDIR)$(bindir)
${INSTALLCMD} -m 755 bin/ctdbd $(DESTDIR)$(sbindir)
${INSTALLCMD} -m 755 bin/smnotify $(DESTDIR)$(bindir)
$(INSTALLCMD) -m 755 bin/ping_pong $(DESTDIR)$(bindir)
@ -223,7 +216,6 @@ install: all
${INSTALLCMD} -m 755 config/events.d/60.nfs $(DESTDIR)$(etcdir)/ctdb/events.d
${INSTALLCMD} -m 755 config/events.d/61.nfstickle $(DESTDIR)$(etcdir)/ctdb/events.d
${INSTALLCMD} -m 755 config/events.d/70.iscsi $(DESTDIR)$(etcdir)/ctdb/events.d
${INSTALLCMD} -m 755 config/events.d/90.ipmux $(DESTDIR)$(etcdir)/ctdb/events.d
${INSTALLCMD} -m 755 config/events.d/91.lvs $(DESTDIR)$(etcdir)/ctdb/events.d
${INSTALLCMD} -m 755 config/events.d/99.routing $(DESTDIR)$(etcdir)/ctdb/events.d
${INSTALLCMD} -m 755 tools/ctdb_diagnostics $(DESTDIR)$(bindir)

View File

@ -58,7 +58,6 @@ CTDB_OPTIONS="$CTDB_OPTIONS --reclock=$CTDB_RECOVERY_LOCK"
[ -z "$CTDB_SOCKET" ] || CTDB_OPTIONS="$CTDB_OPTIONS --socket=$CTDB_SOCKET"
[ -z "$CTDB_PUBLIC_ADDRESSES" ] || CTDB_OPTIONS="$CTDB_OPTIONS --public-addresses=$CTDB_PUBLIC_ADDRESSES"
[ -z "$CTDB_PUBLIC_INTERFACE" ] || CTDB_OPTIONS="$CTDB_OPTIONS --public-interface=$CTDB_PUBLIC_INTERFACE"
[ -z "$CTDB_SINGLE_PUBLIC_IP" ] || CTDB_OPTIONS="$CTDB_OPTIONS --single-public-ip=$CTDB_SINGLE_PUBLIC_IP"
[ -z "$CTDB_DBDIR" ] || CTDB_OPTIONS="$CTDB_OPTIONS --dbdir=$CTDB_DBDIR"
[ -z "$CTDB_DBDIR_PERSISTENT" ] || CTDB_OPTIONS="$CTDB_OPTIONS --dbdir-persistent=$CTDB_DBDIR_PERSISTENT"
[ -z "$CTDB_EVENT_SCRIPT_DIR" ] || CTDB_OPTIONS="$CTDB_OPTIONS --event-script-dir $CTDB_EVENT_SCRIPT_DIR"

View File

@ -29,15 +29,6 @@
# CTDB_LVS_PUBLIC_IP=10.1.1.1
# IPMUX : OBSOLETE use LVS instead
# Should ctdb implement a single public ip address across the entire cluster
# and multiplex incoming connections across the connected nodes
# When using a single public ip you must also specify the public interface!
# This makes all incoming traffic go through a single ctdb node which
# will then forward the packets out acros the other nodes. This will
# impact performance.
# CTDB_SINGLE_PUBLIC_IP=10.1.1.1
# should ctdb manage starting/stopping the Samba service for you?
# default is to not manage Samba
# CTDB_MANAGES_SAMBA=yes

View File

@ -1,87 +0,0 @@
#!/bin/sh
# script to manage the ip multiplexer for a single public address cluster
. $CTDB_BASE/functions
loadconfig ctdb
[ -z "$CTDB_SINGLE_PUBLIC_IP" ] && exit 0
[ -z "$CTDB_PUBLIC_INTERFACE" ] && exit 0
cmd="$1"
shift
PATH=/usr/bin:/bin:/usr/sbin:/sbin:$PATH
case $cmd in
startup)
modprobe ip_queue
ip addr add $CTDB_SINGLE_PUBLIC_IP/32 dev lo scope host >/dev/null 2>/dev/null
# do not respond to ARPs that are for ip addresses with scope 'host'
echo 3 > /proc/sys/net/ipv4/conf/all/arp_ignore
;;
shutdown)
# remove the ip
ip addr del $CTDB_SINGLE_PUBLIC_IP/32 dev lo >/dev/null 2>/dev/null
# remove any iptables rules
iptables -D INPUT -i $CTDB_PUBLIC_INTERFACE -d $CTDB_SINGLE_PUBLIC_IP -j DROP 2> /dev/null
# kill off any ipmux processes
killall -9 ctdb_ipmux >/dev/null 2>/dev/null
# flush our route cache
echo 1 > /proc/sys/net/ipv4/route/flush
;;
takeip)
;;
releaseip)
;;
recovered)
# remove any previous rule for queueing
iptables -D INPUT -d $CTDB_SINGLE_PUBLIC_IP -i $CTDB_PUBLIC_INTERFACE -j QUEUE >/dev/null 2>/dev/null
# kill off any ipmux processes
killall -9 ctdb_ipmux >/dev/null 2>/dev/null
# kill off any tcp connections
iptables -D INPUT -i $CTDB_PUBLIC_INTERFACE -d $CTDB_SINGLE_PUBLIC_IP -j DROP 2> /dev/null
iptables -I INPUT -i $CTDB_PUBLIC_INTERFACE -d $CTDB_SINGLE_PUBLIC_IP -j DROP
kill_tcp_connections $CTDB_SINGLE_PUBLIC_IP
iptables -D INPUT -i $CTDB_PUBLIC_INTERFACE -d $CTDB_SINGLE_PUBLIC_IP -j DROP 2> /dev/null
# are we the recmaster ?
ctdb isnotrecmaster >/dev/null 2>/dev/null || {
# change the ip address to have scope host so we wont respond
# to arps
ip addr del $CTDB_SINGLE_PUBLIC_IP/32 dev lo >/dev/null 2>/dev/null
ip addr add $CTDB_SINGLE_PUBLIC_IP/32 dev lo scope host >/dev/null 2>/dev/null
exit 0
}
# change the scope so we start responding to arps
ip addr del $CTDB_SINGLE_PUBLIC_IP/32 dev lo >/dev/null 2>/dev/null
ip addr add $CTDB_SINGLE_PUBLIC_IP/32 dev lo >/dev/null 2>/dev/null
# send out a gratious arp so our peers will update their arp tables
ctdb gratiousarp $CTDB_SINGLE_PUBLIC_IP $CTDB_PUBLIC_INTERFACE >/dev/null 2>/dev/null
# mark all these for queueing
iptables -I INPUT 1 -d $CTDB_SINGLE_PUBLIC_IP -i $CTDB_PUBLIC_INTERFACE -j QUEUE >/dev/null 2>/dev/null
ctdb_ipmux &
# flush our route cache
echo 1 > /proc/sys/net/ipv4/route/flush
;;
monitor)
;;
esac
exit 0

View File

@ -69,14 +69,6 @@ m4_include(libtdb.m4)
m4_include(libevents.m4)
m4_include(ib/config.m4)
AC_CHECK_HEADERS([libipq.h linux/netfilter.h], [], [], [#include <netinet/in.h>])
IPQ_LIBS=""
if test x"$ac_cv_header_libipq_h" = x"yes"; then
IPQ_LIBS="-lipq"
fi
AC_SUBST(IPQ_LIBS)
AC_CHECK_HEADERS(sched.h)
AC_CHECK_FUNCS(sched_setscheduler)

View File

@ -134,21 +134,6 @@ If you omit this argument when using public addresses or single public ip, ctdb
This is the name of the lock file stored of the shared cluster filesystem that ctdbd uses to arbitrate which node has the role of recovery\-master\. This file must be stored on shared storage\.
.RE
.PP
\-\-single\-public\-ip=<address>
.RS 4
This option is used to activate the "ipmux" or the "lvs" functionality of ctdb where the cluster provides a single public ip address for the entire cluster\. When using this option you must also use the \-\-public\-interface option\.
.sp
In this mode, all nodes of the cluster will expose a single ip address from all nodes with all incoming traffic to the cluster being passed through the current recmaster\. This functionality is similar to using a load\-balancing switch\.
.sp
All incoming packets are sent to the recmaster which will multiplex the clients across all available nodes and pass the packets on to a different node in the cluster to manage the connection based on the clients ip address\. Outgoing packets however are sent directly from the node that was choosen back to the client\. Since all incoming packets are sent through the recmaster this will have a throughput and performance impact when used\. This impact in performance primarily affects write\-performance while read\-performance should be mainly unaffected\. Only use this feature if your environment is mostly\-read (i\.e\. most traffic is from the nodes back to the clients) or if it is not important to get maximum write\-performance to the cluster\.
.sp
This feature is completely controlled from the eventscripts and does not require any CTDBD involvement\. However, the CTDBD daemon does need to know that the "single public ip" exists so that the CTDBD daemon will allow clients to set up killtcp to work on this ip address\.
.sp
CTDBD only allows clients to use killtcp to kill off (RST) tcp connections to/from an ip address that is either a normal public address or to/from the ip address specified by \-\-single\-public\-ip\. No other tcp connections are allowed to be specified with killtcp\.
.sp
Please note that ipmux is obsolete\. Use LVS, not ipmux\. Please see the LVS section in this manpage for instructions on how to configure and use CTDB with LVS\.
.RE
.PP
\-\-socket=<filename>
.RS 4
This specifies the name of the domain socket that ctdbd will create\. This socket is used for local clients to attach to and communicate with the ctdbd daemon\.

View File

@ -84,45 +84,6 @@
</p></dd><dt><span class="term">--reclock=&lt;filename&gt;</span></dt><dd><p>
This is the name of the lock file stored of the shared cluster filesystem that ctdbd uses to arbitrate which node has the role of recovery-master.
This file must be stored on shared storage.
</p></dd><dt><span class="term">--single-public-ip=&lt;address&gt;</span></dt><dd><p>
This option is used to activate the "ipmux" or the "lvs"
functionality of ctdb where the cluster provides a single
public ip address for the entire cluster. When using this option
you must also use the --public-interface option.
</p><p>
In this mode, all nodes of the cluster will expose a single
ip address from all nodes with all incoming traffic to the cluster
being passed through the current recmaster. This functionality
is similar to using a load-balancing switch.
</p><p>
All incoming packets are sent to the recmaster which will multiplex
the clients across all available nodes and pass the packets on to
a different node in the cluster to manage the connection based
on the clients ip address. Outgoing packets however are sent
directly from the node that was choosen back to the client.
Since all incoming packets are sent through the recmaster this will
have a throughput and performance impact when used. This impact
in performance primarily affects write-performance while
read-performance should be mainly unaffected.
Only use this feature if your environment is mostly-read
(i.e. most traffic is from the nodes back to the clients) or
if it is not important to get maximum write-performance to the
cluster.
</p><p>
This feature is completely controlled from the eventscripts and
does not require any CTDBD involvement. However, the CTDBD daemon
does need to know that the "single public ip" exists so that the
CTDBD daemon will allow clients to set up killtcp to work on this
ip address.
</p><p>
CTDBD only allows clients to use killtcp to kill off (RST) tcp
connections to/from an ip address that is either a normal public
address or to/from the ip address specified by --single-public-ip.
No other tcp connections are allowed to be specified with killtcp.
</p><p>
Please note that ipmux is obsolete. Use LVS, not ipmux.
Please see the LVS section in this manpage for instructions on
how to configure and use CTDB with LVS.
</p></dd><dt><span class="term">--socket=&lt;filename&gt;</span></dt><dd><p>
This specifies the name of the domain socket that ctdbd will create. This socket is used for local clients to attach to and communicate with the ctdbd daemon.
</p><p>
@ -154,10 +115,10 @@
implemented in the future.
</p></dd><dt><span class="term">--usage</span></dt><dd><p>
Print useage information to the screen.
</p></dd></dl></div></div><div class="refsect1" lang="en"><a name="id2528846"></a><h2>Private vs Public addresses</h2><p>
</p></dd></dl></div></div><div class="refsect1" lang="en"><a name="id2528779"></a><h2>Private vs Public addresses</h2><p>
When used for ip takeover in a HA environment, each node in a ctdb
cluster has multiple ip addresses assigned to it. One private and one or more public.
</p><div class="refsect2" lang="en"><a name="id2528856"></a><h3>Private address</h3><p>
</p><div class="refsect2" lang="en"><a name="id2528790"></a><h3>Private address</h3><p>
This is the physical ip address of the node which is configured in
linux and attached to a physical interface. This address uniquely
identifies a physical node in the cluster and is the ip addresses
@ -187,7 +148,7 @@
10.1.1.2
10.1.1.3
10.1.1.4
</pre></div><div class="refsect2" lang="en"><a name="id2528904"></a><h3>Public address</h3><p>
</pre></div><div class="refsect2" lang="en"><a name="id2528838"></a><h3>Public address</h3><p>
A public address on the other hand is not attached to an interface.
This address is managed by ctdbd itself and is attached/detached to
a physical node at runtime.
@ -248,7 +209,7 @@
unavailable. 10.1.1.1 can not be failed over to node 2 or node 3 since
these nodes do not have this ip address listed in their public
addresses file.
</p></div></div><div class="refsect1" lang="en"><a name="id2528986"></a><h2>Node status</h2><p>
</p></div></div><div class="refsect1" lang="en"><a name="id2528920"></a><h2>Node status</h2><p>
The current status of each node in the cluster can be viewed by the
'ctdb status' command.
</p><p>
@ -279,50 +240,50 @@
investigated and require an administrative action to rectify. This node
does not perticipate in the CTDB cluster but can still be communicated
with. I.e. ctdb commands can be sent to it.
</p></div><div class="refsect1" lang="en"><a name="id2529049"></a><h2>PUBLIC TUNABLES</h2><p>
</p></div><div class="refsect1" lang="en"><a name="id2528977"></a><h2>PUBLIC TUNABLES</h2><p>
These are the public tuneables that can be used to control how ctdb behaves.
</p><div class="refsect2" lang="en"><a name="id2529059"></a><h3>KeepaliveInterval</h3><p>Default: 1</p><p>
</p><div class="refsect2" lang="en"><a name="id2528988"></a><h3>KeepaliveInterval</h3><p>Default: 1</p><p>
How often should the nodes send keepalives to eachother.
</p></div><div class="refsect2" lang="en"><a name="id2529073"></a><h3>KeepaliveLimit</h3><p>Default: 5</p><p>
</p></div><div class="refsect2" lang="en"><a name="id2529001"></a><h3>KeepaliveLimit</h3><p>Default: 5</p><p>
After how many keepalive intervals without any traffic should a node
wait until marking the peer as DISCONNECTED.
</p></div><div class="refsect2" lang="en"><a name="id2529088"></a><h3>MonitorInterval</h3><p>Default: 15</p><p>
</p></div><div class="refsect2" lang="en"><a name="id2529016"></a><h3>MonitorInterval</h3><p>Default: 15</p><p>
How often should ctdb run the event scripts to check for a nodes health.
</p></div><div class="refsect2" lang="en"><a name="id2529101"></a><h3>TickleUpdateInterval</h3><p>Default: 20</p><p>
</p></div><div class="refsect2" lang="en"><a name="id2529030"></a><h3>TickleUpdateInterval</h3><p>Default: 20</p><p>
How often will ctdb record and store the "tickle" information used to
kickstart stalled tcp connections after a recovery.
</p></div><div class="refsect2" lang="en"><a name="id2529116"></a><h3>EventScriptTimeout</h3><p>Default: 20</p><p>
</p></div><div class="refsect2" lang="en"><a name="id2529044"></a><h3>EventScriptTimeout</h3><p>Default: 20</p><p>
How long should ctdb let an event script run before aborting it and
marking the node unhealthy.
</p></div><div class="refsect2" lang="en"><a name="id2529130"></a><h3>RecoveryBanPeriod</h3><p>Default: 300</p><p>
</p></div><div class="refsect2" lang="en"><a name="id2529059"></a><h3>RecoveryBanPeriod</h3><p>Default: 300</p><p>
If a node becomes banned causing repetitive recovery failures. The node will
eventually become banned from the cluster.
This controls how long the culprit node will be banned from the cluster
before it is allowed to try to join the cluster again.
Dont set to small. A node gets banned for a reason and it is usually due
to real problems with the node.
</p></div><div class="refsect2" lang="en"><a name="id2529149"></a><h3>DatabaseHashSize</h3><p>Default: 100000</p><p>
</p></div><div class="refsect2" lang="en"><a name="id2529078"></a><h3>DatabaseHashSize</h3><p>Default: 100000</p><p>
Size of the hash chains for the local store of the tdbs that ctdb manages.
</p></div><div class="refsect2" lang="en"><a name="id2529163"></a><h3>RerecoveryTimeout</h3><p>Default: 10</p><p>
</p></div><div class="refsect2" lang="en"><a name="id2529092"></a><h3>RerecoveryTimeout</h3><p>Default: 10</p><p>
Once a recovery has completed, no additional recoveries are permitted until this timeout has expired.
</p></div><div class="refsect2" lang="en"><a name="id2529178"></a><h3>EnableBans</h3><p>Default: 1</p><p>
</p></div><div class="refsect2" lang="en"><a name="id2529106"></a><h3>EnableBans</h3><p>Default: 1</p><p>
When set to 0, this disables BANNING completely in the cluster and thus nodes can not get banned, even it they break. Dont set to 0.
</p></div><div class="refsect2" lang="en"><a name="id2529193"></a><h3>DeterministicIPs</h3><p>Default: 1</p><p>
</p></div><div class="refsect2" lang="en"><a name="id2529121"></a><h3>DeterministicIPs</h3><p>Default: 1</p><p>
When enabled, this tunable makes ctdb try to keep public IP addresses locked to specific nodes as far as possible. This makes it easier for debugging since you can know that as long as all nodes are healthy public IP X will always be hosted by node Y.
</p><p>
The cost of using deterministic IP address assignment is that it disables part of the logic where ctdb tries to reduce the number of public IP assignment changes in the cluster. This tunable may increase the number of IP failover/failbacks that are performed on the cluster by a small margin.
</p></div><div class="refsect2" lang="en"><a name="id2529222"></a><h3>DisableWhenUnhealthy</h3><p>Default: 0</p><p>
</p></div><div class="refsect2" lang="en"><a name="id2529146"></a><h3>DisableWhenUnhealthy</h3><p>Default: 0</p><p>
When set, As soon as a node becomes unhealthy, that node will also automatically become permanently DISABLED. Once a node is DISABLED, the only way to make it participate in the cluster again and host services is by manually enabling the node again using 'ctdb enable'.
</p><p>
This disables parts of the resilience and robustness of the cluster and should ONLY be used when the system administrator is actively monitoring the cluster, so that nodes can be enabled again.
</p></div><div class="refsect2" lang="en"><a name="id2529246"></a><h3>NoIPFailback</h3><p>Default: 0</p><p>
</p></div><div class="refsect2" lang="en"><a name="id2529170"></a><h3>NoIPFailback</h3><p>Default: 0</p><p>
When set to 1, ctdb will not perform failback of IP addresses when a node becomes healthy. Ctdb WILL perform failover of public IP addresses when a node becomes UNHEALTHY, but when the node becomes HEALTHY again, ctdb will not fail the addresses back.
</p><p>
Use with caution! Normally when a node becomes available to the cluster
ctdb will try to reassign public IP addresses onto the new node as a way to distribute the workload evenly across the clusternode. Ctdb tries to make sure that all running nodes have approximately the same number of public addresses it hosts.
</p><p>
When you enable this tunable, CTDB will no longer attempt to rebalance the cluster by failing IP addresses back to the new nodes. An unbalanced cluster will therefore remain unbalanced until there is manual intervention from the administrator. When this parameter is set, you can manually fail public IP addresses over to the new node(s) using the 'ctdb moveip' command.
</p></div></div><div class="refsect1" lang="en"><a name="id2529282"></a><h2>LVS</h2><p>
</p></div></div><div class="refsect1" lang="en"><a name="id2529206"></a><h2>LVS</h2><p>
LVS is a mode where CTDB presents one single IP address for the entire
cluster. This is an alternative to using public IP addresses and round-robin
DNS to loadbalance clients across the cluster.
@ -363,7 +324,7 @@ ctdb will try to reassign public IP addresses onto the new node as a way to dist
the processing node back to the clients. For read-intensive i/o patterns you can acheive very high throughput rates in this mode.
</p><p>
Note: you can use LVS and public addresses at the same time.
</p><div class="refsect2" lang="en"><a name="id2529354"></a><h3>Configuration</h3><p>
</p><div class="refsect2" lang="en"><a name="id2529279"></a><h3>Configuration</h3><p>
To activate LVS on a CTDB node you must specify CTDB_PUBLIC_INTERFACE and
CTDB_LVS_PUBLIC_ADDRESS in /etc/sysconfig/ctdb.
</p><p>
@ -386,7 +347,7 @@ You must also specify the "--lvs" command line argument to ctdbd to activete LVS
all of the clients from the node BEFORE you enable LVS. Also make sure
that when you ping these hosts that the traffic is routed out through the
eth0 interface.
</p></div><div class="refsect1" lang="en"><a name="id2529402"></a><h2>REMOTE CLUSTER NODES</h2><p>
</p></div><div class="refsect1" lang="en"><a name="id2529326"></a><h2>REMOTE CLUSTER NODES</h2><p>
It is possible to have a CTDB cluster that spans across a WAN link.
For example where you have a CTDB cluster in your datacentre but you also
want to have one additional CTDB node located at a remote branch site.
@ -415,7 +376,7 @@ CTDB_CAPABILITY_RECMASTER=no
</p><p>
Verify with the command "ctdb getcapabilities" that that node no longer
has the recmaster or the lmaster capabilities.
</p></div><div class="refsect1" lang="en"><a name="id2529453"></a><h2>NAT-GW</h2><p>
</p></div><div class="refsect1" lang="en"><a name="id2529386"></a><h2>NAT-GW</h2><p>
Sometimes it is desireable to run services on the CTDB node which will
need to originate outgoing traffic to external servers. This might
be contacting NIS servers, LDAP servers etc. etc.
@ -438,7 +399,7 @@ CTDB_CAPABILITY_RECMASTER=no
if there are no public addresses assigned to the node.
This is the simplest way but it uses up a lot of ip addresses since you
have to assign both static and also public addresses to each node.
</p><div class="refsect2" lang="en"><a name="id2529493"></a><h3>NAT-GW</h3><p>
</p><div class="refsect2" lang="en"><a name="id2529426"></a><h3>NAT-GW</h3><p>
A second way is to use the built in NAT-GW feature in CTDB.
With NAT-GW you assign one public NATGW address for each natgw group.
Each NATGW group is a set of nodes in the cluster that shares the same
@ -453,7 +414,7 @@ CTDB_CAPABILITY_RECMASTER=no
In each NATGW group, one of the nodes is designated the NAT Gateway
through which all traffic that is originated by nodes in this group
will be routed through if a public addresses are not available.
</p></div><div class="refsect2" lang="en"><a name="id2529523"></a><h3>Configuration</h3><p>
</p></div><div class="refsect2" lang="en"><a name="id2529456"></a><h3>Configuration</h3><p>
NAT-GW is configured in /etc/sysconfigctdb by setting the following
variables:
</p><pre class="screen">
@ -492,31 +453,31 @@ CTDB_CAPABILITY_RECMASTER=no
# CTDB_NATGW_DEFAULT_GATEWAY=10.0.0.1
# CTDB_NATGW_PRIVATE_NETWORK=10.1.1.0/24
# CTDB_NATGW_NODES=/etc/ctdb/natgw_nodes
</pre></div><div class="refsect2" lang="en"><a name="id2476123"></a><h3>CTDB_NATGW_PUBLIC_IP</h3><p>
</pre></div><div class="refsect2" lang="en"><a name="id2529529"></a><h3>CTDB_NATGW_PUBLIC_IP</h3><p>
This is an ip address in the public network that is used for all outgoing
traffic when the public addresses are not assigned.
This address will be assigned to one of the nodes in the cluster which
will masquerade all traffic for the other nodes.
</p><p>
Format of this parameter is IPADDRESS/NETMASK
</p></div><div class="refsect2" lang="en"><a name="id2476141"></a><h3>CTDB_NATGW_PUBLIC_IFACE</h3><p>
</p></div><div class="refsect2" lang="en"><a name="id2529546"></a><h3>CTDB_NATGW_PUBLIC_IFACE</h3><p>
This is the physical interface where the CTDB_NATGW_PUBLIC_IP will be
assigned to. This should be an interface connected to the public network.
</p><p>
Format of this parameter is INTERFACE
</p></div><div class="refsect2" lang="en"><a name="id2476156"></a><h3>CTDB_NATGW_DEFAULT_GATEWAY</h3><p>
</p></div><div class="refsect2" lang="en"><a name="id2529562"></a><h3>CTDB_NATGW_DEFAULT_GATEWAY</h3><p>
This is the default gateway to use on the node that is elected to host
the CTDB_NATGW_PUBLIC_IP. This is the default gateway on the public network.
</p><p>
Format of this parameter is IPADDRESS
</p></div><div class="refsect2" lang="en"><a name="id2476172"></a><h3>CTDB_NATGW_PRIVATE_NETWORK</h3><p>
</p></div><div class="refsect2" lang="en"><a name="id2476133"></a><h3>CTDB_NATGW_PRIVATE_NETWORK</h3><p>
This is the network/netmask used for the interal private network.
</p><p>
Format of this parameter is IPADDRESS/NETMASK
</p></div><div class="refsect2" lang="en"><a name="id2476186"></a><h3>CTDB_NATGW_NODES</h3><p>
</p></div><div class="refsect2" lang="en"><a name="id2476148"></a><h3>CTDB_NATGW_NODES</h3><p>
This is the list of all nodes that belong to the same NATGW group
as this node. The default is /etc/ctdb/natgw_nodes.
</p></div><div class="refsect2" lang="en"><a name="id2476197"></a><h3>Operation</h3><p>
</p></div><div class="refsect2" lang="en"><a name="id2476159"></a><h3>Operation</h3><p>
When the NAT-GW fiunctionality is used, one of the nodes is elected
to act as a NAT router for all the other nodes in the group when
they need to originate traffic to the external public network.
@ -531,11 +492,11 @@ CTDB_CAPABILITY_RECMASTER=no
</p><p>
This is implemented in the 11.natgw eventscript. Please see the
eventscript for further information.
</p></div></div><div class="refsect1" lang="en"><a name="id2476230"></a><h2>ClamAV Daemon</h2><p>
</p></div></div><div class="refsect1" lang="en"><a name="id2476192"></a><h2>ClamAV Daemon</h2><p>
CTDB has support to manage the popular anti-virus daemon ClamAV.
This support is implemented through the
eventscript : /etc/ctdb/events.d/31.clamd.
</p><div class="refsect2" lang="en"><a name="id2476241"></a><h3>Configuration</h3><p>
</p><div class="refsect2" lang="en"><a name="id2476203"></a><h3>Configuration</h3><p>
Start by configuring CLAMAV normally and test that it works. Once this is
done, copy the configuration files over to all the nodes so that all nodes
share identical CLAMAV configurations.
@ -564,10 +525,10 @@ Once you have restarted CTDBD, use
ctdb scriptstatus
</pre><p>
and verify that the 31.clamd eventscript is listed and that it was executed successfully.
</p></div></div><div class="refsect1" lang="en"><a name="id2476302"></a><h2>SEE ALSO</h2><p>
</p></div></div><div class="refsect1" lang="en"><a name="id2476264"></a><h2>SEE ALSO</h2><p>
ctdb(1), onnode(1)
<a class="ulink" href="http://ctdb.samba.org/" target="_top">http://ctdb.samba.org/</a>
</p></div><div class="refsect1" lang="en"><a name="id2476315"></a><h2>COPYRIGHT/LICENSE</h2><div class="literallayout"><p><br>
</p></div><div class="refsect1" lang="en"><a name="id2476277"></a><h2>COPYRIGHT/LICENSE</h2><div class="literallayout"><p><br>
Copyright (C) Andrew Tridgell 2007<br>
Copyright (C) Ronnie sahlberg 2007<br>
<br>

View File

@ -254,56 +254,6 @@
</listitem>
</varlistentry>
<varlistentry><term>--single-public-ip=&lt;address&gt;</term>
<listitem>
<para>
This option is used to activate the "ipmux" or the "lvs"
functionality of ctdb where the cluster provides a single
public ip address for the entire cluster. When using this option
you must also use the --public-interface option.
</para>
<para>
In this mode, all nodes of the cluster will expose a single
ip address from all nodes with all incoming traffic to the cluster
being passed through the current recmaster. This functionality
is similar to using a load-balancing switch.
</para>
<para>
All incoming packets are sent to the recmaster which will multiplex
the clients across all available nodes and pass the packets on to
a different node in the cluster to manage the connection based
on the clients ip address. Outgoing packets however are sent
directly from the node that was choosen back to the client.
Since all incoming packets are sent through the recmaster this will
have a throughput and performance impact when used. This impact
in performance primarily affects write-performance while
read-performance should be mainly unaffected.
Only use this feature if your environment is mostly-read
(i.e. most traffic is from the nodes back to the clients) or
if it is not important to get maximum write-performance to the
cluster.
</para>
<para>
This feature is completely controlled from the eventscripts and
does not require any CTDBD involvement. However, the CTDBD daemon
does need to know that the "single public ip" exists so that the
CTDBD daemon will allow clients to set up killtcp to work on this
ip address.
</para>
<para>
CTDBD only allows clients to use killtcp to kill off (RST) tcp
connections to/from an ip address that is either a normal public
address or to/from the ip address specified by --single-public-ip.
No other tcp connections are allowed to be specified with killtcp.
</para>
<para>
Please note that ipmux is obsolete. Use LVS, not ipmux.
Please see the LVS section in this manpage for instructions on
how to configure and use CTDB with LVS.
</para>
</listitem>
</varlistentry>
<varlistentry><term>--socket=&lt;filename&gt;</term>
<listitem>
<para>

View File

@ -114,7 +114,6 @@ fi
%{_sysconfdir}/ctdb/events.d/60.nfs
%{_sysconfdir}/ctdb/events.d/61.nfstickle
%{_sysconfdir}/ctdb/events.d/70.iscsi
%{_sysconfdir}/ctdb/events.d/90.ipmux
%{_sysconfdir}/ctdb/events.d/91.lvs
%{_sysconfdir}/ctdb/events.d/99.routing
%{_sysconfdir}/ctdb/statd-callout
@ -122,7 +121,6 @@ fi
%{_bindir}/ctdb
%{_bindir}/smnotify
%{_bindir}/ping_pong
%{_bindir}/ctdb_ipmux
%{_bindir}/ctdb_diagnostics
%{_bindir}/onnode
%{_mandir}/man1/ctdb.1.gz

View File

@ -1,268 +0,0 @@
/*
simple ip multiplexer
Copyright (C) Ronnie Sahlberg 2007
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, see <http://www.gnu.org/licenses/>.
*/
#include "includes.h"
#include "lib/events/events.h"
#include "system/filesys.h"
#include "system/network.h"
#include "popt.h"
#include "cmdline.h"
#include "ctdb.h"
#include "ctdb_private.h"
#if defined(HAVE_LINUX_NETFILTER_H) && defined(HAVE_LIBIPQ_H)
#include <linux/netfilter.h>
#include <libipq.h>
#define CONTROL_TIMEOUT() timeval_current_ofs(5, 0)
struct ipmux_node {
uint32_t pnn;
ctdb_sock_addr addr;
};
struct ipmux_node *ipmux_nodes;
/*
This function is used to open a raw socket to send tickles from
*/
int ctdb_sys_open_sending_socket(void)
{
int s, ret;
uint32_t one = 1;
s = socket(AF_INET, SOCK_RAW, htons(IPPROTO_RAW));
if (s == -1) {
DEBUG(DEBUG_CRIT,(__location__ " failed to open raw socket (%s)\n",
strerror(errno)));
return -1;
}
ret = setsockopt(s, SOL_IP, IP_HDRINCL, &one, sizeof(one));
if (ret != 0) {
DEBUG(DEBUG_CRIT,(__location__ " failed to setup IP headers (%s)\n",
strerror(errno)));
close(s);
return -1;
}
set_nonblocking(s);
set_close_on_exec(s);
return s;
}
/*
main program
*/
int main(int argc, const char *argv[])
{
struct ctdb_context *ctdb;
struct poptOption popt_options[] = {
POPT_AUTOHELP
POPT_CTDB_CMDLINE
POPT_TABLEEND
};
int opt;
const char **extra_argv;
int extra_argc = 0;
int ret;
poptContext pc;
struct event_context *ev;
uint32_t mypnn, recmaster;
TALLOC_CTX *mem_ctx=NULL;
struct ctdb_node_map *nodemap;
int i, num_nodes;
int s;
struct ipq_handle *ipqh;
#define PKTSIZE 65535
unsigned char pktbuf[PKTSIZE];
ipq_packet_msg_t *ipqp;
struct iphdr *ip;
int hash;
pc = poptGetContext(argv[0], argc, argv, popt_options, POPT_CONTEXT_KEEP_FIRST);
while ((opt = poptGetNextOpt(pc)) != -1) {
switch (opt) {
default:
fprintf(stderr, "Invalid option %s: %s\n",
poptBadOption(pc, 0), poptStrerror(opt));
exit(1);
}
}
/* talloc_enable_leak_report_full(); */
/* setup the remaining options for the main program to use */
extra_argv = poptGetArgs(pc);
if (extra_argv) {
extra_argv++;
while (extra_argv[extra_argc]) extra_argc++;
}
ev = event_context_init(NULL);
ctdb = ctdb_cmdline_client(ev);
mem_ctx = talloc_new(ctdb);
/* get our pnn */
mypnn = ctdb_ctrl_getpnn(ctdb, CONTROL_TIMEOUT(), CTDB_CURRENT_NODE);
if (mypnn == (uint32_t)-1) {
DEBUG(0,("IPMUX: Failed to get local pnn - exiting\n"));
talloc_free(mem_ctx);
exit(10);
}
/* get the recmaster */
ret = ctdb_ctrl_getrecmaster(ctdb, mem_ctx, CONTROL_TIMEOUT(), CTDB_CURRENT_NODE, &recmaster);
if (ret != 0) {
DEBUG(0,("IPMUX: Failed to get recmaster - exiting\n"));
talloc_free(mem_ctx);
exit(10);
}
/* verify we are the recmaster */
if (recmaster != mypnn) {
DEBUG(0,("IPMUX: we are not the recmaster - exiting\n"));
talloc_free(mem_ctx);
exit(10);
}
/* get the list of nodes */
ret = ctdb_ctrl_getnodemap(ctdb, CONTROL_TIMEOUT(), CTDB_CURRENT_NODE, mem_ctx, &nodemap);
if (ret != 0) {
DEBUG(0,("IPMUX: failed to get the nodemap - exiting\n"));
talloc_free(mem_ctx);
exit(10);
}
/* count how many connected nodes we have */
num_nodes = 0;
for (i=0; i<nodemap->num; i++) {
if (nodemap->nodes[i].flags & NODE_FLAGS_DISCONNECTED) {
continue;
}
num_nodes++;
}
if (num_nodes == 0) {
DEBUG(0,("IPMUX: no connected nodes - exiting\n"));
talloc_free(mem_ctx);
exit(10);
}
ipmux_nodes = talloc_array(mem_ctx, struct ipmux_node, num_nodes);
if (ipmux_nodes == NULL) {
DEBUG(0,("IPMUX: failed to allocate ipmux node array - exiting\n"));
talloc_free(mem_ctx);
exit(10);
}
/* populate the ipmux node array */
num_nodes = 0;
for (i=0; i<nodemap->num; i++) {
if (nodemap->nodes[i].flags & NODE_FLAGS_DISCONNECTED) {
continue;
}
ipmux_nodes[num_nodes].pnn = i;
ipmux_nodes[num_nodes].addr = nodemap->nodes[i].addr;
num_nodes++;
}
/* open a raw socket to send the packets out through */
s = ctdb_sys_open_sending_socket();
if (s == -1) {
DEBUG(0,("IPMUX: failed to open raw socket - exiting\n"));
talloc_free(mem_ctx);
exit(10);
}
/* open the ipq handle */
ipqh = ipq_create_handle(0, PF_INET);
if (ipqh == NULL) {
DEBUG(0,("IPMUX: failed to create ipq handle - exiting\n"));
talloc_free(mem_ctx);
exit(10);
}
ret = ipq_set_mode(ipqh, IPQ_COPY_PACKET, PKTSIZE);
if (ret < 0) {
DEBUG(0,("IPMUX: failed to set ipq mode. make sure the ip_queue module is loaded - exiting\n"));
talloc_free(mem_ctx);
exit(10);
}
while (1) {
/* wait for the next packet */
ret = ipq_read(ipqh, pktbuf, PKTSIZE, 0);
if (ret <= 0) {
continue;
}
/* read the packet */
ipqp = ipq_get_packet(pktbuf);
if (ipqp == NULL) {
continue;
}
/* calculate a hash based on the clients ip address */
ip = (struct iphdr *)&ipqp->payload[0];
/* ntohl here since the client ip addresses are much more
likely to differ in the lower bits than the hight bits */
hash = ntohl(ip->saddr) % num_nodes;
/* if the packet is hashed to the current host, then
just accept it and let the kernel pass it onto
the local stack
*/
if (ipmux_nodes[hash].pnn == mypnn) {
ipq_set_verdict(ipqh, ipqp->packet_id, NF_ACCEPT, 0, pktbuf);
continue;
}
/* we have hashed it to one of the other nodes, so
send the packet off and tell the kernel to not worry
about this packet any more
*/
ret = sendto(s, &ipqp->payload[0], ipqp->data_len, 0, (struct sockaddr_in *)&ipmux_nodes[hash].addr, sizeof(ctdb_sock_addr));
ipq_set_verdict(ipqh, ipqp->packet_id, NF_DROP, 0, pktbuf);
}
return 0;
}
#else
int main(void)
{
printf("ipmux tool disabled - lacking netfilter and libipq development libs\n");
return 1;
}
#endif