dejan at suse
May 14, 2012, 9:11 AM
Post #12 of 19
(722 views)
Permalink
|
Hi Yves,
On Fri, May 11, 2012 at 11:27:23AM -0400, Yves Trudeau wrote:
> Hi Dejan,
>
> Le 2012-05-11 11:04, Dejan Muhamedagic a écrit :
>> Hi,
>>
>> On Fri, May 11, 2012 at 10:25:19AM -0400, Yves Trudeau wrote:
>>> Hi Dejan,
>>> ok, here the latest version using
>>> ${INSTANCE_ATTR_NAME}_MYSQL_MASTER_IP
>>
>> I'd vote for less "yelling" as cluster configurations are mostly
>> lower case.
>
> :) Ok, never thought about the impact of the upper case...
>
>>
>>> and I agree with the merits of
>>> this :) I added a paragraph explaining the use of the attribute in
>>> the longdesc of the meta-data.
>>
>> Excellent! Looks good to me. If nobody objects, we can push this
>> come Monday.
>>
>> BTW, on what repository is this based? Can you produce a set of
>> patches to be applied to upstream?
>
> set of patches... here's _one_ patch bringing the latest commit of
> ClusterLabs/resource-agents to my version. I also lowered the case of
> the attribute.
Good!
One thing I forgot though: Can you also provide a commit
message, i.e. a patch description (what changed, what's new,
etc).
Cheers,
Dejan
> Regards,
>
> Yves
>
>
>
>
>
>>
>> Cheers,
>>
>> Dejan
>>
>>
>>
>>
>>> Regards,
>>>
>>> Yves
>>>
>>>
>>> Le 2012-05-11 09:21, Dejan Muhamedagic a écrit :
>>>> Hi Yves,
>>>>
>>>> On Fri, May 11, 2012 at 08:45:06AM -0400, Yves Trudeau wrote:
>>>>> Hi Dejan,
>>>>> I changed the name of the attribute to REPL_MASTER_IP
>>>>
>>>> Let's quote from one of previous emails:
>>>>
>>>> 2. Is it possible/plausible to have more than one mysql
>>>> instance? If so, then the attribute name should include the
>>>> instance name. Say ${INSTANCE_NAME}_mysql_replication_IP or
>>>> something to that extent. Also, it would make for a better
>>>> looking configuration. "IP" doesn't really say much.
>>>>
>>>>> and added 2
>>>>> lines of comment for the get_local_ip function.
>>>>
>>>> Again, a quote:
>>>>
>>>> 3. This attribute is part of the configuration and supposed to
>>>> be setup by the user. Please document that in the meta-data.
>>>>
>>>> Note that one of the key words here is "meta-data." That is
>>>> supposed to be documentation for the users, not for developers.
>>>> Users don't normally read the code.
>>>>
>>>>> Is that inline with
>>>>> what you want?
>>>>
>>>> OT:
>>>>
>>>> It doesn't really matter what _I_ want. We're having a discussion
>>>> here on how to improve the feature. It is just by chance that I am
>>>> right now the only one talking about it.
>>>>
>>>> Cheers,
>>>>
>>>> Dejan
>>>>
>>>>> Regards,
>>>>>
>>>>> Yves
>>>>>
>>>>> Le 2012-05-11 04:45, Dejan Muhamedagic a écrit :
>>>>>> Hi Yves,
>>>>>>
>>>>>> It would be good not to start a new thread for the same
>>>>>> discussion.
>>>>>>
>>>>>> On Thu, May 10, 2012 at 05:06:25PM -0400, Yves Trudeau wrote:
>>>>>>> Hi Dejan,
>>>>>>> here's another modified patch for the mysql agent of the commit
>>>>>>> version 4c18035 (git [at] github:y-trudeau/resource-agents.git branch
>>>>>>> mysql-repl). This patch implements fallback on uname -n if the node
>>>>>>> IP attribute is not present and uses the instance name for the
>>>>>>> replication info attribute.
>>>>>>
>>>>>> Hmm, it looks like there was a misunderstanding here. The
>>>>>> attribute named "IP" is still named "IP" :)
>>>>>>
>>>>>> And we're still missing the documentation for that attribute.
>>>>>>
>>>>>>> I am also working with Raoul to get me
>>>>>>> back on track with git.
>>>>>>
>>>>>> Good!
>>>>>>
>>>>>> Cheers,
>>>>>>
>>>>>> Dejan
>>>>>>
>>>>>>>
>>>>>>> Regards,
>>>>>>>
>>>>>>> Yves
>>>>>>
>>>>>>> --- mysql.4c18035b7 2012-05-04 16:23:45.238421759 -0400
>>>>>>> +++ resource-agents-prm/heartbeat/mysql 2012-05-10 11:01:02.538421042 -0400
>>>>>>> @@ -109,7 +109,7 @@
>>>>>>> : ${OCF_RESKEY_max_slave_lag=${OCF_RESKEY_max_slave_lag_default}}
>>>>>>> : ${OCF_RESKEY_evict_outdated_slaves=${OCF_RESKEY_evict_outdated_slaves_default}}
>>>>>>>
>>>>>>> -: ${OCF_RESKEY_reader_attribute=${OCF_RESKEY_evict_reader_attribute_default}}
>>>>>>> +: ${OCF_RESKEY_reader_attribute=${OCF_RESKEY_reader_attribute_default}}
>>>>>>>
>>>>>>> #######################################################################
>>>>>>>
>>>>>>> @@ -328,7 +328,7 @@
>>>>>>> </longdesc>
>>>>>>> <shortdesc lang="en">Sets the node attribute that determines
>>>>>>> whether a node is usable for clients to read from.</shortdesc>
>>>>>>> -<content type="boolean" default="${OCF_RESKEY_reader_attribute_default}" />
>>>>>>> +<content type="string" default="${OCF_RESKEY_reader_attribute_default}" />
>>>>>>> </parameter>
>>>>>>> </parameters>
>>>>>>>
>>>>>>> @@ -355,11 +355,13 @@
>>>>>>> MYSQL_OPTIONS_LOCAL="-S $OCF_RESKEY_socket --connect_timeout=10"
>>>>>>> MYSQL_OPTIONS_REPL="$MYSQL_OPTIONS_LOCAL --user=$OCF_RESKEY_replication_user --password=$OCF_RESKEY_replication_passwd"
>>>>>>> MYSQL_OPTIONS_TEST="$MYSQL_OPTIONS_LOCAL --user=$OCF_RESKEY_test_user --password=$OCF_RESKEY_test_passwd"
>>>>>>> +MYSQL_TOO_MANY_CONN_ERR=1040
>>>>>>>
>>>>>>> CRM_MASTER="${HA_SBIN_DIR}/crm_master -l reboot "
>>>>>>> HOSTNAME=`uname -n`
>>>>>>> CRM_ATTR="${HA_SBIN_DIR}/crm_attribute -N $HOSTNAME "
>>>>>>> INSTANCE_ATTR_NAME=`echo ${OCF_RESOURCE_INSTANCE}| awk -F : '{print $1}'`
>>>>>>> +CRM_ATTR_REPL_INFO="${HA_SBIN_DIR}/crm_attribute --type crm_config --name ${INSTANCE_ATTR_NAME}_REPL_INFO -s mysql_replication"
>>>>>>>
>>>>>>> #######################################################################
>>>>>>>
>>>>>>> @@ -461,14 +463,14 @@
>>>>>>>
>>>>>>> check_slave() {
>>>>>>> # Checks slave status
>>>>>>> - local rc new_master_IP
>>>>>>> + local rc new_master
>>>>>>>
>>>>>>> get_slave_info
>>>>>>> rc=$?
>>>>>>>
>>>>>>> if [ $rc -eq 0 ]; then
>>>>>>> # Did we receive an error other than max_connections?
>>>>>>> - if [ $last_errno -ne 0 -a $last_errno -ne 1040 ]; then
>>>>>>> + if [ $last_errno -ne 0 -a $last_errno -ne "$MYSQL_TOO_MANY_CONN_ERR" ]; then
>>>>>>> # Whoa. Replication ran into an error. This slave has
>>>>>>> # diverged from its master. Make sure this resource
>>>>>>> # doesn't restart in place.
>>>>>>> @@ -484,7 +486,7 @@
>>>>>>> fi
>>>>>>>
>>>>>>> # If we got max_connections, let's remove the vip
>>>>>>> - if [ $last_errno -eq 1040 ]; then
>>>>>>> + if [ $last_errno -eq "$MYSQL_TOO_MANY_CONN_ERR" ]; then
>>>>>>> set_reader_attr 0
>>>>>>> exit $OCF_SUCCESS
>>>>>>> fi
>>>>>>> @@ -496,9 +498,9 @@
>>>>>>> ocf_log warn "MySQL Slave IO threads currently not running."
>>>>>>>
>>>>>>> # Sanity check, are we at least on the right master
>>>>>>> - new_master_IP=`${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication --query -q | cut -d'|' -f1`
>>>>>>> + new_master=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f1`
>>>>>>>
>>>>>>> - if [ "$master_host" != "$new_master_IP" ]; then
>>>>>>> + if [ "$master_host" != "$new_master" ]; then
>>>>>>> # Not pointing to the right master, not good, removing the VIPs
>>>>>>> set_reader_attr 0
>>>>>>>
>>>>>>> @@ -570,23 +572,23 @@
>>>>>>> }
>>>>>>>
>>>>>>> set_master() {
>>>>>>> - local new_master_IP master_log_file master_log_pos
>>>>>>> + local new_master master_log_file master_log_pos
>>>>>>> local master_params
>>>>>>>
>>>>>>> - new_master_IP=`${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication --query -q | cut -d'|' -f1`
>>>>>>> + new_master=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f1`
>>>>>>>
>>>>>>> # Keep replication position
>>>>>>> get_slave_info
>>>>>>>
>>>>>>> - if [ "$master_log_file" -a "$new_master_IP" = "$master_host" ]; then
>>>>>>> + if [ "$master_log_file" -a "$new_master" = "$master_host" ]; then
>>>>>>> # master_params=", MASTER_LOG_FILE='$master_log_file', \
>>>>>>> # MASTER_LOG_POS=$master_log_pos"
>>>>>>> ocf_log info "Kept master pos for $master_host : $master_log_file:$master_log_pos"
>>>>>>> rm -f $tmpfile
>>>>>>> return
>>>>>>> else
>>>>>>> - master_log_file=`${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication --query -q | cut -d'|' -f2`
>>>>>>> - master_log_pos=`${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication --query -q | cut -d'|' -f3`
>>>>>>> + master_log_file=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f2`
>>>>>>> + master_log_pos=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f3`
>>>>>>> if [ -n "$master_log_file" -a -n "$master_log_pos" ]; then
>>>>>>> master_params=", MASTER_LOG_FILE='$master_log_file', \
>>>>>>> MASTER_LOG_POS=$master_log_pos"
>>>>>>> @@ -601,7 +603,7 @@
>>>>>>> # reset with RESET MASTER.
>>>>>>>
>>>>>>> ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
>>>>>>> - -e "CHANGE MASTER TO MASTER_HOST='$new_master_IP', \
>>>>>>> + -e "CHANGE MASTER TO MASTER_HOST='$new_master', \
>>>>>>> MASTER_USER='$OCF_RESKEY_replication_user', \
>>>>>>> MASTER_PASSWORD='$OCF_RESKEY_replication_passwd' $master_params"
>>>>>>> rm -f $tmpfile
>>>>>>> @@ -628,15 +630,15 @@
>>>>>>> while true; do
>>>>>>> $MYSQL $MYSQL_OPTIONS_REPL \
>>>>>>> -e 'SHOW PROCESSLIST\G'> $tmpfile
>>>>>>> - if grep 'Waiting for master to send event' $tmpfile>/dev/null; then
>>>>>>> + if grep -i 'Waiting for master to send event' $tmpfile>/dev/null; then
>>>>>>> ocf_log info "MySQL slave has finished reading master binary log"
>>>>>>> break
>>>>>>> fi
>>>>>>> - if grep 'Reconnecting after a failed master event read' $tmpfile>/dev/null; then
>>>>>>> + if grep -i 'Reconnecting after a failed master event read' $tmpfile>/dev/null; then
>>>>>>> ocf_log info "Master is down, no more binary logs to come"
>>>>>>> break
>>>>>>> fi
>>>>>>> - if grep 'Connecting to master' $tmpfile>/dev/null; then
>>>>>>> + if grep -i 'Connecting to master' $tmpfile>/dev/null; then
>>>>>>> ocf_log info "Master is down, no more binary logs to come"
>>>>>>> break
>>>>>>> fi
>>>>>>> @@ -660,7 +662,7 @@
>>>>>>> while true; do
>>>>>>> $MYSQL $MYSQL_OPTIONS_REPL \
>>>>>>> -e 'SHOW PROCESSLIST\G'> $tmpfile
>>>>>>> - if grep 'Has read all relay log' $tmpfile>/dev/null; then
>>>>>>> + if grep -i 'Has read all relay log' $tmpfile>/dev/null; then
>>>>>>> ocf_log info "MySQL slave has finished processing relay log"
>>>>>>> break
>>>>>>> fi
>>>>>>> @@ -744,8 +746,15 @@
>>>>>>> # Determines what IP address is attached to the current host. The output of the
>>>>>>> # crm_attribute command looks like this:
>>>>>>> # scope=nodes name=IP value=10.2.2.161
>>>>>>> +# If the IP node attribute is not defined, fallback is to uname -n
>>>>>>> get_local_ip() {
>>>>>>> - $CRM_ATTR -l forever -n IP -q -G
>>>>>>> + local IP
>>>>>>> + IP=`$CRM_ATTR -l forever -n IP -q -G`
>>>>>>> + if [ ! $? -eq 0 ]; then
>>>>>>> + uname -n
>>>>>>> + else
>>>>>>> + echo $IP
>>>>>>> + fi
>>>>>>> }
>>>>>>>
>>>>>>> #######################################################################
>>>>>>> @@ -783,7 +792,7 @@
>>>>>>>
>>>>>>> mysql_status() {
>>>>>>> if [ ! -e $OCF_RESKEY_pid ]; then
>>>>>>> - ocf_log err "MySQL is not running"
>>>>>>> + ocf_log $1 "MySQL is not running"
>>>>>>> return $OCF_NOT_RUNNING;
>>>>>>> fi
>>>>>>>
>>>>>>> @@ -797,7 +806,7 @@
>>>>>>> if [ $? -eq 0 ]; then
>>>>>>> return $OCF_SUCCESS;
>>>>>>> else
>>>>>>> - ocf_log err "MySQL not running: removing old PID file"
>>>>>>> + ocf_log $1 "MySQL not running: removing old PID file"
>>>>>>> rm -f $OCF_RESKEY_pid
>>>>>>> return $OCF_NOT_RUNNING;
>>>>>>> fi
>>>>>>> @@ -811,8 +820,9 @@
>>>>>>> if ocf_is_probe; then
>>>>>>> status_loglevel="info"
>>>>>>> fi
>>>>>>> -
>>>>>>> +
>>>>>>> mysql_status $status_loglevel
>>>>>>> +
>>>>>>> rc=$?
>>>>>>>
>>>>>>> # TODO: check max connections error
>>>>>>> @@ -856,7 +866,7 @@
>>>>>>> set_reader_attr 0
>>>>>>> fi
>>>>>>>
>>>>>>> - mysql_status
>>>>>>> + mysql_status info
>>>>>>> if [ $? = $OCF_SUCCESS ]; then
>>>>>>> ocf_log info "MySQL already running"
>>>>>>> return $OCF_SUCCESS
>>>>>>> @@ -930,7 +940,7 @@
>>>>>>> # Let the CRM/LRM time us out if required.
>>>>>>> start_wait=1
>>>>>>> while [ $start_wait = 1 ]; do
>>>>>>> - mysql_status
>>>>>>> + mysql_status info
>>>>>>> rc=$?
>>>>>>> if [ $rc = $OCF_SUCCESS ]; then
>>>>>>> start_wait=0
>>>>>>> @@ -1019,7 +1029,7 @@
>>>>>>> count=0
>>>>>>> while [ $count -lt $shutdown_timeout ]
>>>>>>> do
>>>>>>> - mysql_status
>>>>>>> + mysql_status info
>>>>>>> rc=$?
>>>>>>> if [ $rc = $OCF_NOT_RUNNING ]; then
>>>>>>> break
>>>>>>> @@ -1029,7 +1039,7 @@
>>>>>>> ocf_log debug "MySQL still hasn't stopped yet. Waiting..."
>>>>>>> done
>>>>>>>
>>>>>>> - mysql_status
>>>>>>> + mysql_status info
>>>>>>> if [ $? != $OCF_NOT_RUNNING ]; then
>>>>>>> ocf_log info "MySQL failed to stop after ${shutdown_timeout}s using SIGTERM. Trying SIGKILL..."
>>>>>>> /bin/kill -KILL $pid> /dev/null
>>>>>>> @@ -1044,7 +1054,7 @@
>>>>>>> mysql_promote() {
>>>>>>> local master_info
>>>>>>>
>>>>>>> - if ( ! mysql_status ); then
>>>>>>> + if ( ! mysql_status err ); then
>>>>>>> return $OCF_NOT_RUNNING
>>>>>>> fi
>>>>>>> ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
>>>>>>> @@ -1053,7 +1063,7 @@
>>>>>>> # Set Master Info in CIB, cluster level attribute
>>>>>>> update_data_master_status
>>>>>>> master_info="$(get_local_ip)|$(get_master_status File)|$(get_master_status Position)"
>>>>>>> - ${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication -v "$master_info"
>>>>>>> + ${CRM_ATTR_REPL_INFO} -v "$master_info"
>>>>>>> rm -f $tmpfile
>>>>>>>
>>>>>>> set_read_only off || return $OCF_ERR_GENERIC
>>>>>>> @@ -1070,7 +1080,7 @@
>>>>>>> }
>>>>>>>
>>>>>>> mysql_demote() {
>>>>>>> - if ! mysql_status; then
>>>>>>> + if ! mysql_status err; then
>>>>>>> return $OCF_NOT_RUNNING
>>>>>>> fi
>>>>>>>
>>>>>>> @@ -1177,6 +1187,7 @@
>>>>>>> # The log directory must be a directory owned by root, with permissions 0700,
>>>>>>> # and the log must be writable and not a symlink.
>>>>>>> ##########################################################################
>>>>>>> +DEBUG_LOG="/tmp/mysql.ocf.ra.debug/log"
>>>>>>> if [ "${DEBUG_LOG}" -a -w "${DEBUG_LOG}" -a ! -L "${DEBUG_LOG}" ]; then
>>>>>>> DEBUG_LOG_DIR="${DEBUG_LOG%/*}"
>>>>>>> if [ -d "${DEBUG_LOG_DIR}" ]; then
>>>>>>> @@ -1214,7 +1225,7 @@
>>>>>>> case "$1" in
>>>>>>> start) mysql_start;;
>>>>>>> stop) mysql_stop;;
>>>>>>> - status) mysql_status;;
>>>>>>> + status) mysql_status err;;
>>>>>>> monitor) mysql_monitor;;
>>>>>>> promote) mysql_promote;;
>>>>>>> demote) mysql_demote;;
>>>>>>
>>>>>>> _______________________________________________________
>>>>>>> Linux-HA-Dev: Linux-HA-Dev [at] lists
>>>>>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
>>>>>>> Home Page: http://linux-ha.org/
>>>>>>
>>>>>> _______________________________________________________
>>>>>> Linux-HA-Dev: Linux-HA-Dev [at] lists
>>>>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
>>>>>> Home Page: http://linux-ha.org/
>>>>>>
>>>>
>>>>> --- mysql.4c18035b7 2012-05-04 16:23:45.238421759 -0400
>>>>> +++ resource-agents-prm/heartbeat/mysql 2012-05-11 08:41:20.868420414 -0400
>>>>> @@ -109,7 +109,7 @@
>>>>> : ${OCF_RESKEY_max_slave_lag=${OCF_RESKEY_max_slave_lag_default}}
>>>>> : ${OCF_RESKEY_evict_outdated_slaves=${OCF_RESKEY_evict_outdated_slaves_default}}
>>>>>
>>>>> -: ${OCF_RESKEY_reader_attribute=${OCF_RESKEY_evict_reader_attribute_default}}
>>>>> +: ${OCF_RESKEY_reader_attribute=${OCF_RESKEY_reader_attribute_default}}
>>>>>
>>>>> #######################################################################
>>>>>
>>>>> @@ -328,7 +328,7 @@
>>>>> </longdesc>
>>>>> <shortdesc lang="en">Sets the node attribute that determines
>>>>> whether a node is usable for clients to read from.</shortdesc>
>>>>> -<content type="boolean" default="${OCF_RESKEY_reader_attribute_default}" />
>>>>> +<content type="string" default="${OCF_RESKEY_reader_attribute_default}" />
>>>>> </parameter>
>>>>> </parameters>
>>>>>
>>>>> @@ -355,11 +355,13 @@
>>>>> MYSQL_OPTIONS_LOCAL="-S $OCF_RESKEY_socket --connect_timeout=10"
>>>>> MYSQL_OPTIONS_REPL="$MYSQL_OPTIONS_LOCAL --user=$OCF_RESKEY_replication_user --password=$OCF_RESKEY_replication_passwd"
>>>>> MYSQL_OPTIONS_TEST="$MYSQL_OPTIONS_LOCAL --user=$OCF_RESKEY_test_user --password=$OCF_RESKEY_test_passwd"
>>>>> +MYSQL_TOO_MANY_CONN_ERR=1040
>>>>>
>>>>> CRM_MASTER="${HA_SBIN_DIR}/crm_master -l reboot "
>>>>> HOSTNAME=`uname -n`
>>>>> CRM_ATTR="${HA_SBIN_DIR}/crm_attribute -N $HOSTNAME "
>>>>> INSTANCE_ATTR_NAME=`echo ${OCF_RESOURCE_INSTANCE}| awk -F : '{print $1}'`
>>>>> +CRM_ATTR_REPL_INFO="${HA_SBIN_DIR}/crm_attribute --type crm_config --name ${INSTANCE_ATTR_NAME}_REPL_INFO -s mysql_replication"
>>>>>
>>>>> #######################################################################
>>>>>
>>>>> @@ -461,14 +463,14 @@
>>>>>
>>>>> check_slave() {
>>>>> # Checks slave status
>>>>> - local rc new_master_IP
>>>>> + local rc new_master
>>>>>
>>>>> get_slave_info
>>>>> rc=$?
>>>>>
>>>>> if [ $rc -eq 0 ]; then
>>>>> # Did we receive an error other than max_connections?
>>>>> - if [ $last_errno -ne 0 -a $last_errno -ne 1040 ]; then
>>>>> + if [ $last_errno -ne 0 -a $last_errno -ne "$MYSQL_TOO_MANY_CONN_ERR" ]; then
>>>>> # Whoa. Replication ran into an error. This slave has
>>>>> # diverged from its master. Make sure this resource
>>>>> # doesn't restart in place.
>>>>> @@ -484,7 +486,7 @@
>>>>> fi
>>>>>
>>>>> # If we got max_connections, let's remove the vip
>>>>> - if [ $last_errno -eq 1040 ]; then
>>>>> + if [ $last_errno -eq "$MYSQL_TOO_MANY_CONN_ERR" ]; then
>>>>> set_reader_attr 0
>>>>> exit $OCF_SUCCESS
>>>>> fi
>>>>> @@ -496,9 +498,9 @@
>>>>> ocf_log warn "MySQL Slave IO threads currently not running."
>>>>>
>>>>> # Sanity check, are we at least on the right master
>>>>> - new_master_IP=`${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication --query -q | cut -d'|' -f1`
>>>>> + new_master=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f1`
>>>>>
>>>>> - if [ "$master_host" != "$new_master_IP" ]; then
>>>>> + if [ "$master_host" != "$new_master" ]; then
>>>>> # Not pointing to the right master, not good, removing the VIPs
>>>>> set_reader_attr 0
>>>>>
>>>>> @@ -570,23 +572,23 @@
>>>>> }
>>>>>
>>>>> set_master() {
>>>>> - local new_master_IP master_log_file master_log_pos
>>>>> + local new_master master_log_file master_log_pos
>>>>> local master_params
>>>>>
>>>>> - new_master_IP=`${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication --query -q | cut -d'|' -f1`
>>>>> + new_master=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f1`
>>>>>
>>>>> # Keep replication position
>>>>> get_slave_info
>>>>>
>>>>> - if [ "$master_log_file" -a "$new_master_IP" = "$master_host" ]; then
>>>>> + if [ "$master_log_file" -a "$new_master" = "$master_host" ]; then
>>>>> # master_params=", MASTER_LOG_FILE='$master_log_file', \
>>>>> # MASTER_LOG_POS=$master_log_pos"
>>>>> ocf_log info "Kept master pos for $master_host : $master_log_file:$master_log_pos"
>>>>> rm -f $tmpfile
>>>>> return
>>>>> else
>>>>> - master_log_file=`${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication --query -q | cut -d'|' -f2`
>>>>> - master_log_pos=`${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication --query -q | cut -d'|' -f3`
>>>>> + master_log_file=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f2`
>>>>> + master_log_pos=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f3`
>>>>> if [ -n "$master_log_file" -a -n "$master_log_pos" ]; then
>>>>> master_params=", MASTER_LOG_FILE='$master_log_file', \
>>>>> MASTER_LOG_POS=$master_log_pos"
>>>>> @@ -601,7 +603,7 @@
>>>>> # reset with RESET MASTER.
>>>>>
>>>>> ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
>>>>> - -e "CHANGE MASTER TO MASTER_HOST='$new_master_IP', \
>>>>> + -e "CHANGE MASTER TO MASTER_HOST='$new_master', \
>>>>> MASTER_USER='$OCF_RESKEY_replication_user', \
>>>>> MASTER_PASSWORD='$OCF_RESKEY_replication_passwd' $master_params"
>>>>> rm -f $tmpfile
>>>>> @@ -628,15 +630,15 @@
>>>>> while true; do
>>>>> $MYSQL $MYSQL_OPTIONS_REPL \
>>>>> -e 'SHOW PROCESSLIST\G'> $tmpfile
>>>>> - if grep 'Waiting for master to send event' $tmpfile>/dev/null; then
>>>>> + if grep -i 'Waiting for master to send event' $tmpfile>/dev/null; then
>>>>> ocf_log info "MySQL slave has finished reading master binary log"
>>>>> break
>>>>> fi
>>>>> - if grep 'Reconnecting after a failed master event read' $tmpfile>/dev/null; then
>>>>> + if grep -i 'Reconnecting after a failed master event read' $tmpfile>/dev/null; then
>>>>> ocf_log info "Master is down, no more binary logs to come"
>>>>> break
>>>>> fi
>>>>> - if grep 'Connecting to master' $tmpfile>/dev/null; then
>>>>> + if grep -i 'Connecting to master' $tmpfile>/dev/null; then
>>>>> ocf_log info "Master is down, no more binary logs to come"
>>>>> break
>>>>> fi
>>>>> @@ -660,7 +662,7 @@
>>>>> while true; do
>>>>> $MYSQL $MYSQL_OPTIONS_REPL \
>>>>> -e 'SHOW PROCESSLIST\G'> $tmpfile
>>>>> - if grep 'Has read all relay log' $tmpfile>/dev/null; then
>>>>> + if grep -i 'Has read all relay log' $tmpfile>/dev/null; then
>>>>> ocf_log info "MySQL slave has finished processing relay log"
>>>>> break
>>>>> fi
>>>>> @@ -744,8 +746,17 @@
>>>>> # Determines what IP address is attached to the current host. The output of the
>>>>> # crm_attribute command looks like this:
>>>>> # scope=nodes name=IP value=10.2.2.161
>>>>> +# If the REPL_MASTER_IP node attribute is not defined, fallback is to uname -n
>>>>> +# The REPL_MASTER_IP is the IP address that will be used for the
>>>>> +# change master to command.
>>>>> get_local_ip() {
>>>>> - $CRM_ATTR -l forever -n IP -q -G
>>>>> + local IP
>>>>> + IP=`$CRM_ATTR -l forever -n REPL_MASTER_IP -q -G`
>>>>> + if [ ! $? -eq 0 ]; then
>>>>> + uname -n
>>>>> + else
>>>>> + echo $IP
>>>>> + fi
>>>>> }
>>>>>
>>>>> #######################################################################
>>>>> @@ -783,7 +794,7 @@
>>>>>
>>>>> mysql_status() {
>>>>> if [ ! -e $OCF_RESKEY_pid ]; then
>>>>> - ocf_log err "MySQL is not running"
>>>>> + ocf_log $1 "MySQL is not running"
>>>>> return $OCF_NOT_RUNNING;
>>>>> fi
>>>>>
>>>>> @@ -797,7 +808,7 @@
>>>>> if [ $? -eq 0 ]; then
>>>>> return $OCF_SUCCESS;
>>>>> else
>>>>> - ocf_log err "MySQL not running: removing old PID file"
>>>>> + ocf_log $1 "MySQL not running: removing old PID file"
>>>>> rm -f $OCF_RESKEY_pid
>>>>> return $OCF_NOT_RUNNING;
>>>>> fi
>>>>> @@ -811,8 +822,9 @@
>>>>> if ocf_is_probe; then
>>>>> status_loglevel="info"
>>>>> fi
>>>>> -
>>>>> +
>>>>> mysql_status $status_loglevel
>>>>> +
>>>>> rc=$?
>>>>>
>>>>> # TODO: check max connections error
>>>>> @@ -856,7 +868,7 @@
>>>>> set_reader_attr 0
>>>>> fi
>>>>>
>>>>> - mysql_status
>>>>> + mysql_status info
>>>>> if [ $? = $OCF_SUCCESS ]; then
>>>>> ocf_log info "MySQL already running"
>>>>> return $OCF_SUCCESS
>>>>> @@ -930,7 +942,7 @@
>>>>> # Let the CRM/LRM time us out if required.
>>>>> start_wait=1
>>>>> while [ $start_wait = 1 ]; do
>>>>> - mysql_status
>>>>> + mysql_status info
>>>>> rc=$?
>>>>> if [ $rc = $OCF_SUCCESS ]; then
>>>>> start_wait=0
>>>>> @@ -1019,7 +1031,7 @@
>>>>> count=0
>>>>> while [ $count -lt $shutdown_timeout ]
>>>>> do
>>>>> - mysql_status
>>>>> + mysql_status info
>>>>> rc=$?
>>>>> if [ $rc = $OCF_NOT_RUNNING ]; then
>>>>> break
>>>>> @@ -1029,7 +1041,7 @@
>>>>> ocf_log debug "MySQL still hasn't stopped yet. Waiting..."
>>>>> done
>>>>>
>>>>> - mysql_status
>>>>> + mysql_status info
>>>>> if [ $? != $OCF_NOT_RUNNING ]; then
>>>>> ocf_log info "MySQL failed to stop after ${shutdown_timeout}s using SIGTERM. Trying SIGKILL..."
>>>>> /bin/kill -KILL $pid> /dev/null
>>>>> @@ -1044,7 +1056,7 @@
>>>>> mysql_promote() {
>>>>> local master_info
>>>>>
>>>>> - if ( ! mysql_status ); then
>>>>> + if ( ! mysql_status err ); then
>>>>> return $OCF_NOT_RUNNING
>>>>> fi
>>>>> ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
>>>>> @@ -1053,7 +1065,7 @@
>>>>> # Set Master Info in CIB, cluster level attribute
>>>>> update_data_master_status
>>>>> master_info="$(get_local_ip)|$(get_master_status File)|$(get_master_status Position)"
>>>>> - ${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication -v "$master_info"
>>>>> + ${CRM_ATTR_REPL_INFO} -v "$master_info"
>>>>> rm -f $tmpfile
>>>>>
>>>>> set_read_only off || return $OCF_ERR_GENERIC
>>>>> @@ -1070,7 +1082,7 @@
>>>>> }
>>>>>
>>>>> mysql_demote() {
>>>>> - if ! mysql_status; then
>>>>> + if ! mysql_status err; then
>>>>> return $OCF_NOT_RUNNING
>>>>> fi
>>>>>
>>>>> @@ -1177,6 +1189,7 @@
>>>>> # The log directory must be a directory owned by root, with permissions 0700,
>>>>> # and the log must be writable and not a symlink.
>>>>> ##########################################################################
>>>>> +DEBUG_LOG="/tmp/mysql.ocf.ra.debug/log"
>>>>> if [ "${DEBUG_LOG}" -a -w "${DEBUG_LOG}" -a ! -L "${DEBUG_LOG}" ]; then
>>>>> DEBUG_LOG_DIR="${DEBUG_LOG%/*}"
>>>>> if [ -d "${DEBUG_LOG_DIR}" ]; then
>>>>> @@ -1214,7 +1227,7 @@
>>>>> case "$1" in
>>>>> start) mysql_start;;
>>>>> stop) mysql_stop;;
>>>>> - status) mysql_status;;
>>>>> + status) mysql_status err;;
>>>>> monitor) mysql_monitor;;
>>>>> promote) mysql_promote;;
>>>>> demote) mysql_demote;;
>>>>
>>>>> _______________________________________________________
>>>>> Linux-HA-Dev: Linux-HA-Dev [at] lists
>>>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
>>>>> Home Page: http://linux-ha.org/
>>>>
>>>> _______________________________________________________
>>>> Linux-HA-Dev: Linux-HA-Dev [at] lists
>>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
>>>> Home Page: http://linux-ha.org/
>>>>
>>
>>> --- mysql.4c18035b7 2012-05-04 16:23:45.238421759 -0400
>>> +++ resource-agents-prm/heartbeat/mysql 2012-05-11 10:14:28.148420134 -0400
>>> @@ -109,7 +109,22 @@
>>> : ${OCF_RESKEY_max_slave_lag=${OCF_RESKEY_max_slave_lag_default}}
>>> : ${OCF_RESKEY_evict_outdated_slaves=${OCF_RESKEY_evict_outdated_slaves_default}}
>>>
>>> -: ${OCF_RESKEY_reader_attribute=${OCF_RESKEY_evict_reader_attribute_default}}
>>> +: ${OCF_RESKEY_reader_attribute=${OCF_RESKEY_reader_attribute_default}}
>>> +
>>> +#######################################################################
>>> +# Convenience variables
>>> +
>>> +MYSQL=$OCF_RESKEY_client_binary
>>> +MYSQL_OPTIONS_LOCAL="-S $OCF_RESKEY_socket --connect_timeout=10"
>>> +MYSQL_OPTIONS_REPL="$MYSQL_OPTIONS_LOCAL --user=$OCF_RESKEY_replication_user --password=$OCF_RESKEY_replication_passwd"
>>> +MYSQL_OPTIONS_TEST="$MYSQL_OPTIONS_LOCAL --user=$OCF_RESKEY_test_user --password=$OCF_RESKEY_test_passwd"
>>> +MYSQL_TOO_MANY_CONN_ERR=1040
>>> +
>>> +CRM_MASTER="${HA_SBIN_DIR}/crm_master -l reboot "
>>> +HOSTNAME=`uname -n`
>>> +CRM_ATTR="${HA_SBIN_DIR}/crm_attribute -N $HOSTNAME "
>>> +INSTANCE_ATTR_NAME=`echo ${OCF_RESOURCE_INSTANCE}| awk -F : '{print $1}'`
>>> +CRM_ATTR_REPL_INFO="${HA_SBIN_DIR}/crm_attribute --type crm_config --name ${INSTANCE_ATTR_NAME}_REPL_INFO -s mysql_replication"
>>>
>>> #######################################################################
>>>
>>> @@ -141,6 +156,13 @@
>>> Resource script for MySQL.
>>> May manage a standalone MySQL database, a clone set with externally
>>> managed replication, or a complete master/slave replication setup.
>>> +
>>> +While managing replication, the default behavior is to use uname -n
>>> +values in the change master to command. Other IPs can be specified
>>> +manually by adding a node attribute \${INSTANCE_ATTR_NAME}_MYSQL_MASTER_IP
>>> +giving the IP to use for replication. For example, if the mysql primitive
>>> +you are using is p_mysql, the attribute to set will be
>>> +p_mysql_MYSQL_MASTER_IP.
>>> </longdesc>
>>> <shortdesc lang="en">Manages a MySQL database instance</shortdesc>
>>> <parameters>
>>> @@ -328,7 +350,7 @@
>>> </longdesc>
>>> <shortdesc lang="en">Sets the node attribute that determines
>>> whether a node is usable for clients to read from.</shortdesc>
>>> -<content type="boolean" default="${OCF_RESKEY_reader_attribute_default}" />
>>> +<content type="string" default="${OCF_RESKEY_reader_attribute_default}" />
>>> </parameter>
>>> </parameters>
>>>
>>> @@ -348,20 +370,6 @@
>>> </resource-agent>
>>> END
>>> }
>>> -#######################################################################
>>> -# Convenience variables
>>> -
>>> -MYSQL=$OCF_RESKEY_client_binary
>>> -MYSQL_OPTIONS_LOCAL="-S $OCF_RESKEY_socket --connect_timeout=10"
>>> -MYSQL_OPTIONS_REPL="$MYSQL_OPTIONS_LOCAL --user=$OCF_RESKEY_replication_user --password=$OCF_RESKEY_replication_passwd"
>>> -MYSQL_OPTIONS_TEST="$MYSQL_OPTIONS_LOCAL --user=$OCF_RESKEY_test_user --password=$OCF_RESKEY_test_passwd"
>>> -
>>> -CRM_MASTER="${HA_SBIN_DIR}/crm_master -l reboot "
>>> -HOSTNAME=`uname -n`
>>> -CRM_ATTR="${HA_SBIN_DIR}/crm_attribute -N $HOSTNAME "
>>> -INSTANCE_ATTR_NAME=`echo ${OCF_RESOURCE_INSTANCE}| awk -F : '{print $1}'`
>>> -
>>> -#######################################################################
>>>
>>> # Convenience functions
>>>
>>> @@ -461,14 +469,14 @@
>>>
>>> check_slave() {
>>> # Checks slave status
>>> - local rc new_master_IP
>>> + local rc new_master
>>>
>>> get_slave_info
>>> rc=$?
>>>
>>> if [ $rc -eq 0 ]; then
>>> # Did we receive an error other than max_connections?
>>> - if [ $last_errno -ne 0 -a $last_errno -ne 1040 ]; then
>>> + if [ $last_errno -ne 0 -a $last_errno -ne "$MYSQL_TOO_MANY_CONN_ERR" ]; then
>>> # Whoa. Replication ran into an error. This slave has
>>> # diverged from its master. Make sure this resource
>>> # doesn't restart in place.
>>> @@ -484,7 +492,7 @@
>>> fi
>>>
>>> # If we got max_connections, let's remove the vip
>>> - if [ $last_errno -eq 1040 ]; then
>>> + if [ $last_errno -eq "$MYSQL_TOO_MANY_CONN_ERR" ]; then
>>> set_reader_attr 0
>>> exit $OCF_SUCCESS
>>> fi
>>> @@ -496,9 +504,9 @@
>>> ocf_log warn "MySQL Slave IO threads currently not running."
>>>
>>> # Sanity check, are we at least on the right master
>>> - new_master_IP=`${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication --query -q | cut -d'|' -f1`
>>> + new_master=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f1`
>>>
>>> - if [ "$master_host" != "$new_master_IP" ]; then
>>> + if [ "$master_host" != "$new_master" ]; then
>>> # Not pointing to the right master, not good, removing the VIPs
>>> set_reader_attr 0
>>>
>>> @@ -570,23 +578,23 @@
>>> }
>>>
>>> set_master() {
>>> - local new_master_IP master_log_file master_log_pos
>>> + local new_master master_log_file master_log_pos
>>> local master_params
>>>
>>> - new_master_IP=`${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication --query -q | cut -d'|' -f1`
>>> + new_master=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f1`
>>>
>>> # Keep replication position
>>> get_slave_info
>>>
>>> - if [ "$master_log_file" -a "$new_master_IP" = "$master_host" ]; then
>>> + if [ "$master_log_file" -a "$new_master" = "$master_host" ]; then
>>> # master_params=", MASTER_LOG_FILE='$master_log_file', \
>>> # MASTER_LOG_POS=$master_log_pos"
>>> ocf_log info "Kept master pos for $master_host : $master_log_file:$master_log_pos"
>>> rm -f $tmpfile
>>> return
>>> else
>>> - master_log_file=`${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication --query -q | cut -d'|' -f2`
>>> - master_log_pos=`${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication --query -q | cut -d'|' -f3`
>>> + master_log_file=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f2`
>>> + master_log_pos=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f3`
>>> if [ -n "$master_log_file" -a -n "$master_log_pos" ]; then
>>> master_params=", MASTER_LOG_FILE='$master_log_file', \
>>> MASTER_LOG_POS=$master_log_pos"
>>> @@ -601,7 +609,7 @@
>>> # reset with RESET MASTER.
>>>
>>> ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
>>> - -e "CHANGE MASTER TO MASTER_HOST='$new_master_IP', \
>>> + -e "CHANGE MASTER TO MASTER_HOST='$new_master', \
>>> MASTER_USER='$OCF_RESKEY_replication_user', \
>>> MASTER_PASSWORD='$OCF_RESKEY_replication_passwd' $master_params"
>>> rm -f $tmpfile
>>> @@ -628,15 +636,15 @@
>>> while true; do
>>> $MYSQL $MYSQL_OPTIONS_REPL \
>>> -e 'SHOW PROCESSLIST\G'> $tmpfile
>>> - if grep 'Waiting for master to send event' $tmpfile>/dev/null; then
>>> + if grep -i 'Waiting for master to send event' $tmpfile>/dev/null; then
>>> ocf_log info "MySQL slave has finished reading master binary log"
>>> break
>>> fi
>>> - if grep 'Reconnecting after a failed master event read' $tmpfile>/dev/null; then
>>> + if grep -i 'Reconnecting after a failed master event read' $tmpfile>/dev/null; then
>>> ocf_log info "Master is down, no more binary logs to come"
>>> break
>>> fi
>>> - if grep 'Connecting to master' $tmpfile>/dev/null; then
>>> + if grep -i 'Connecting to master' $tmpfile>/dev/null; then
>>> ocf_log info "Master is down, no more binary logs to come"
>>> break
>>> fi
>>> @@ -660,7 +668,7 @@
>>> while true; do
>>> $MYSQL $MYSQL_OPTIONS_REPL \
>>> -e 'SHOW PROCESSLIST\G'> $tmpfile
>>> - if grep 'Has read all relay log' $tmpfile>/dev/null; then
>>> + if grep -i 'Has read all relay log' $tmpfile>/dev/null; then
>>> ocf_log info "MySQL slave has finished processing relay log"
>>> break
>>> fi
>>> @@ -744,8 +752,17 @@
>>> # Determines what IP address is attached to the current host. The output of the
>>> # crm_attribute command looks like this:
>>> # scope=nodes name=IP value=10.2.2.161
>>> +# If the ${INSTANCE_ATTR_NAME}_MYSQL_MASTER_IP node attribute is not defined, fallback is to uname -n
>>> +# The ${INSTANCE_ATTR_NAME}_MYSQL_MASTER_IP is the IP address that will be used for the
>>> +# change master to command.
>>> get_local_ip() {
>>> - $CRM_ATTR -l forever -n IP -q -G
>>> + local IP
>>> + IP=`$CRM_ATTR -l forever -n ${INSTANCE_ATTR_NAME}_MYSQL_MASTER_IP -q -G`
>>> + if [ ! $? -eq 0 ]; then
>>> + uname -n
>>> + else
>>> + echo $IP
>>> + fi
>>> }
>>>
>>> #######################################################################
>>> @@ -783,7 +800,7 @@
>>>
>>> mysql_status() {
>>> if [ ! -e $OCF_RESKEY_pid ]; then
>>> - ocf_log err "MySQL is not running"
>>> + ocf_log $1 "MySQL is not running"
>>> return $OCF_NOT_RUNNING;
>>> fi
>>>
>>> @@ -797,7 +814,7 @@
>>> if [ $? -eq 0 ]; then
>>> return $OCF_SUCCESS;
>>> else
>>> - ocf_log err "MySQL not running: removing old PID file"
>>> + ocf_log $1 "MySQL not running: removing old PID file"
>>> rm -f $OCF_RESKEY_pid
>>> return $OCF_NOT_RUNNING;
>>> fi
>>> @@ -811,8 +828,9 @@
>>> if ocf_is_probe; then
>>> status_loglevel="info"
>>> fi
>>> -
>>> +
>>> mysql_status $status_loglevel
>>> +
>>> rc=$?
>>>
>>> # TODO: check max connections error
>>> @@ -856,7 +874,7 @@
>>> set_reader_attr 0
>>> fi
>>>
>>> - mysql_status
>>> + mysql_status info
>>> if [ $? = $OCF_SUCCESS ]; then
>>> ocf_log info "MySQL already running"
>>> return $OCF_SUCCESS
>>> @@ -930,7 +948,7 @@
>>> # Let the CRM/LRM time us out if required.
>>> start_wait=1
>>> while [ $start_wait = 1 ]; do
>>> - mysql_status
>>> + mysql_status info
>>> rc=$?
>>> if [ $rc = $OCF_SUCCESS ]; then
>>> start_wait=0
>>> @@ -1019,7 +1037,7 @@
>>> count=0
>>> while [ $count -lt $shutdown_timeout ]
>>> do
>>> - mysql_status
>>> + mysql_status info
>>> rc=$?
>>> if [ $rc = $OCF_NOT_RUNNING ]; then
>>> break
>>> @@ -1029,7 +1047,7 @@
>>> ocf_log debug "MySQL still hasn't stopped yet. Waiting..."
>>> done
>>>
>>> - mysql_status
>>> + mysql_status info
>>> if [ $? != $OCF_NOT_RUNNING ]; then
>>> ocf_log info "MySQL failed to stop after ${shutdown_timeout}s using SIGTERM. Trying SIGKILL..."
>>> /bin/kill -KILL $pid> /dev/null
>>> @@ -1044,7 +1062,7 @@
>>> mysql_promote() {
>>> local master_info
>>>
>>> - if ( ! mysql_status ); then
>>> + if ( ! mysql_status err ); then
>>> return $OCF_NOT_RUNNING
>>> fi
>>> ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
>>> @@ -1053,7 +1071,7 @@
>>> # Set Master Info in CIB, cluster level attribute
>>> update_data_master_status
>>> master_info="$(get_local_ip)|$(get_master_status File)|$(get_master_status Position)"
>>> - ${HA_SBIN_DIR}/crm_attribute --type crm_config --name replication_info -s mysql_replication -v "$master_info"
>>> + ${CRM_ATTR_REPL_INFO} -v "$master_info"
>>> rm -f $tmpfile
>>>
>>> set_read_only off || return $OCF_ERR_GENERIC
>>> @@ -1070,7 +1088,7 @@
>>> }
>>>
>>> mysql_demote() {
>>> - if ! mysql_status; then
>>> + if ! mysql_status err; then
>>> return $OCF_NOT_RUNNING
>>> fi
>>>
>>> @@ -1177,6 +1195,7 @@
>>> # The log directory must be a directory owned by root, with permissions 0700,
>>> # and the log must be writable and not a symlink.
>>> ##########################################################################
>>> +DEBUG_LOG="/tmp/mysql.ocf.ra.debug/log"
>>> if [ "${DEBUG_LOG}" -a -w "${DEBUG_LOG}" -a ! -L "${DEBUG_LOG}" ]; then
>>> DEBUG_LOG_DIR="${DEBUG_LOG%/*}"
>>> if [ -d "${DEBUG_LOG_DIR}" ]; then
>>> @@ -1214,7 +1233,7 @@
>>> case "$1" in
>>> start) mysql_start;;
>>> stop) mysql_stop;;
>>> - status) mysql_status;;
>>> + status) mysql_status err;;
>>> monitor) mysql_monitor;;
>>> promote) mysql_promote;;
>>> demote) mysql_demote;;
>>
>>> _______________________________________________________
>>> Linux-HA-Dev: Linux-HA-Dev [at] lists
>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
>>> Home Page: http://linux-ha.org/
>>
>> _______________________________________________________
>> Linux-HA-Dev: Linux-HA-Dev [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
>> Home Page: http://linux-ha.org/
>>
> --- mysql.latest 2012-05-11 11:21:05.848420091 -0400
> +++ resource-agents-prm/heartbeat/mysql 2012-05-11 11:19:33.088420104 -0400
> @@ -11,6 +11,7 @@
> # Sebastian Reitenbach: add OpenBSD defaults, more cleanup
> # Narayan Newton: add Gentoo/Debian defaults
> # Marian Marinov, Florian Haas: add replication capability
> +# Yves Trudeau, Baron Schwartz: add VIP support and improve replication
> #
> # Support: linux-ha [at] lists
> # License: GNU General Public License (GPL)
> @@ -38,6 +39,7 @@
> # OCF_RESKEY_log
> # OCF_RESKEY_pid
> # OCF_RESKEY_socket
> +
> #######################################################################
> # Initialization:
>
> @@ -76,6 +78,7 @@
> OCF_RESKEY_replication_port_default="3306"
> OCF_RESKEY_max_slave_lag_default="3600"
> OCF_RESKEY_evict_outdated_slaves_default="false"
> +OCF_RESKEY_reader_attribute_default="readable"
>
> : ${OCF_RESKEY_binary=${OCF_RESKEY_binary_default}}
> MYSQL_BINDIR=`dirname ${OCF_RESKEY_binary}`
> @@ -106,6 +109,23 @@
> : ${OCF_RESKEY_max_slave_lag=${OCF_RESKEY_max_slave_lag_default}}
> : ${OCF_RESKEY_evict_outdated_slaves=${OCF_RESKEY_evict_outdated_slaves_default}}
>
> +: ${OCF_RESKEY_reader_attribute=${OCF_RESKEY_reader_attribute_default}}
> +
> +#######################################################################
> +# Convenience variables
> +
> +MYSQL=$OCF_RESKEY_client_binary
> +MYSQL_OPTIONS_LOCAL="-S $OCF_RESKEY_socket --connect_timeout=10"
> +MYSQL_OPTIONS_REPL="$MYSQL_OPTIONS_LOCAL --user=$OCF_RESKEY_replication_user --password=$OCF_RESKEY_replication_passwd"
> +MYSQL_OPTIONS_TEST="$MYSQL_OPTIONS_LOCAL --user=$OCF_RESKEY_test_user --password=$OCF_RESKEY_test_passwd"
> +MYSQL_TOO_MANY_CONN_ERR=1040
> +
> +CRM_MASTER="${HA_SBIN_DIR}/crm_master -l reboot "
> +HOSTNAME=`uname -n`
> +CRM_ATTR="${HA_SBIN_DIR}/crm_attribute -N $HOSTNAME "
> +INSTANCE_ATTR_NAME=`echo ${OCF_RESOURCE_INSTANCE}| awk -F : '{print $1}'`
> +CRM_ATTR_REPL_INFO="${HA_SBIN_DIR}/crm_attribute --type crm_config --name ${INSTANCE_ATTR_NAME}_REPL_INFO -s mysql_replication"
> +
> #######################################################################
>
> usage() {
> @@ -133,9 +153,16 @@
> <version>1.0</version>
>
> <longdesc lang="en">
> -Resource script for MySQL.
> +Resource script for MySQL.
> May manage a standalone MySQL database, a clone set with externally
> managed replication, or a complete master/slave replication setup.
> +
> +While managing replication, the default behavior is to use uname -n
> +values in the change master to command. Other IPs can be specified
> +manually by adding a node attribute \${INSTANCE_ATTR_NAME}_mysql_master_IP
> +giving the IP to use for replication. For example, if the mysql primitive
> +you are using is p_mysql, the attribute to set will be
> +p_mysql_mysql_master_IP.
> </longdesc>
> <shortdesc lang="en">Manages a MySQL database instance</shortdesc>
> <parameters>
> @@ -222,7 +249,7 @@
>
> <parameter name="test_user" unique="0" required="0">
> <longdesc lang="en">
> -MySQL test user
> +MySQL test user, must have select privilege on test_table
> </longdesc>
> <shortdesc lang="en">MySQL test user</shortdesc>
> <content type="string" default="${OCF_RESKEY_test_user_default}" />
> @@ -259,7 +286,8 @@
> MySQL replication, for setting and resetting the master host, and for
> setting and unsetting read-only mode. Because of that, this user must
> have SUPER, REPLICATION SLAVE, REPLICATION CLIENT, and PROCESS
> -privileges on all nodes within the cluster.
> +privileges on all nodes within the cluster. Mandatory if you define
> +a master-slave resource.
> </longdesc>
> <shortdesc lang="en">MySQL replication user</shortdesc>
> <content type="string" default="${OCF_RESKEY_replication_user_default}" />
> @@ -268,6 +296,7 @@
> <parameter name="replication_passwd" unique="0" required="0">
> <longdesc lang="en">
> MySQL replication password. Used for replication client and slave.
> +Mandatory if you define a master-slave resource.
> </longdesc>
> <shortdesc lang="en">MySQL replication user password</shortdesc>
> <content type="string" default="${OCF_RESKEY_replication_passwd_default}" />
> @@ -307,6 +336,22 @@
> <content type="boolean" default="${OCF_RESKEY_evict_outdated_slaves_default}" />
> </parameter>
>
> +<parameter name="reader_attribute" unique="1" required="0">
> +<longdesc lang="en">
> +An attribute that the RA can manage to specify whether a node
> +can be read from. This node attribute will be 1 if it's fine to
> +read from the node, and 0 otherwise (for example, when a slave
> +has lagged too far behind the master).
> +
> +A typical example for the use of this attribute would be to tie
> +a set of IP addresses to MySQL slaves that can be read from.
> +
> +This parameter is only meaningful in master/slave set configurations.
> +</longdesc>
> +<shortdesc lang="en">Sets the node attribute that determines
> +whether a node is usable for clients to read from.</shortdesc>
> +<content type="string" default="${OCF_RESKEY_reader_attribute_default}" />
> +</parameter>
> </parameters>
>
> <actions>
> @@ -325,19 +370,6 @@
> </resource-agent>
> END
> }
> -#######################################################################
> -# Convenience variables
> -
> -MYSQL=$OCF_RESKEY_client_binary
> -MYSQL_OPTIONS_LOCAL="-S $OCF_RESKEY_socket --connect_timeout=10"
> -MYSQL_OPTIONS_REPL="--user=$OCF_RESKEY_replication_user --password=$OCF_RESKEY_replication_passwd"
> -
> -CRM_MASTER="${HA_SBIN_DIR}/crm_master -l reboot "
> -HOSTNAME=`uname -n`
> -CRM_ATTR="${HA_SBIN_DIR}/crm_attribute -N $HOSTNAME -l forever"
> -INSTANCE_ATTR_NAME=`echo ${OCF_RESOURCE_INSTANCE}| awk -F : '{print $1}'`
> -
> -#######################################################################
>
> # Convenience functions
>
> @@ -354,28 +386,17 @@
> else
> ro_val="off"
> fi
> - local mysql_options
> - mysql_options="$MYSQL_OPTIONS_LOCAL"
> - if [ -n $OCF_RESKEY_replication_user ]; then
> - mysql_options="$mysql_options $MYSQL_OPTIONS_REPL"
> - fi
> - ocf_run $MYSQL $mysql_options \
> + ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
> -e "SET GLOBAL read_only=${ro_val}"
> }
>
> get_read_only() {
> # Check if read-only is set
> - local mysql_options
> local read_only_state
>
> - mysql_options="$MYSQL_OPTIONS_LOCAL"
> - if [ -n $OCF_RESKEY_replication_user ]; then
> - mysql_options="$mysql_options $MYSQL_OPTIONS_REPL"
> - fi
> -
> - read_only_state=`$MYSQL $mysql_options \
> + read_only_state=`$MYSQL $MYSQL_OPTIONS_REPL \
> -e "SHOW VARIABLES" | grep read_only | awk '{print $2}'`
> -
> +
> if [ "$read_only_state" = "ON" ]; then
> return 0
> else
> @@ -389,7 +410,6 @@
> # SLAVE STATUS creates an empty result set, 0 otherwise.
> local rc
> local tmpfile
> - local mysql_options
>
> # Check whether this machine should be slave
> if ! ocf_is_ms || ! get_read_only; then
> @@ -398,10 +418,8 @@
>
> tmpfile=`mktemp ${HA_RSCTMP}/is_slave.${OCF_RESOURCE_INSTANCE}.XXXXXX`
>
> - mysql_options="$MYSQL_OPTIONS_LOCAL $MYSQL_OPTIONS_REPL"
> -
> - $MYSQL $mysql_options \
> - -e 'SHOW SLAVE STATUS\G' > $tmpfile
> + $MYSQL $MYSQL_OPTIONS_REPL \
> + -e 'SHOW SLAVE STATUS\G' > $tmpfile
>
> # "SHOW SLAVE STATUS" returns an empty set if instance is not a
> # replication slave
> @@ -425,10 +443,8 @@
>
> tmpfile=`mktemp ${HA_RSCTMP}/check_slave.${OCF_RESOURCE_INSTANCE}.XXXXXX`
>
> - mysql_options="$MYSQL_OPTIONS_LOCAL $MYSQL_OPTIONS_REPL"
> -
> - $MYSQL $mysql_options \
> - -e 'SHOW SLAVE STATUS\G' > $tmpfile
> + $MYSQL $MYSQL_OPTIONS_REPL \
> + -e 'SHOW SLAVE STATUS\G' > $tmpfile
>
> if [ -s $tmpfile ]; then
> master_host=`parse_slave_info Master_Host $tmpfile`
> @@ -440,7 +456,6 @@
> slave_io=`parse_slave_info Slave_IO_Running $tmpfile`
> last_errno=`parse_slave_info Last_Errno $tmpfile`
> secs_behind=`parse_slave_info Seconds_Behind_Master $tmpfile`
> -
> ocf_log debug "MySQL instance running as a replication slave"
> else
> # Instance produced an empty "SHOW SLAVE STATUS" output --
> @@ -454,19 +469,32 @@
>
> check_slave() {
> # Checks slave status
> - local rc
> + local rc new_master
>
> get_slave_info
> rc=$?
>
> if [ $rc -eq 0 ]; then
> - if [ $last_errno -ne 0 ]; then
> + # Did we receive an error other than max_connections?
> + if [ $last_errno -ne 0 -a $last_errno -ne "$MYSQL_TOO_MANY_CONN_ERR" ]; then
> # Whoa. Replication ran into an error. This slave has
> # diverged from its master. Make sure this resource
> # doesn't restart in place.
> ocf_log err "MySQL instance configured for replication, but replication has failed."
> ocf_log err "See $tmpfile for details"
> - exit $OCF_ERR_INSTALLED
> +
> + # Just pull the reader VIP away, killing MySQL here would be pretty evil
> + # on a loaded server
> +
> + set_reader_attr 0
> + exit $OCF_SUCCESS
> +
> + fi
> +
> + # If we got max_connections, let's remove the vip
> + if [ $last_errno -eq "$MYSQL_TOO_MANY_CONN_ERR" ]; then
> + set_reader_attr 0
> + exit $OCF_SUCCESS
> fi
>
> if [ "$slave_io" != 'Yes' ]; then
> @@ -474,15 +502,35 @@
> # temporarily shut down, and the slave may just be
> # reconnecting. A warning can't hurt, though.
> ocf_log warn "MySQL Slave IO threads currently not running."
> +
> + # Sanity check, are we at least on the right master
> + new_master=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f1`
> +
> + if [ "$master_host" != "$new_master" ]; then
> + # Not pointing to the right master, not good, removing the VIPs
> + set_reader_attr 0
> +
> + exit $OCF_SUCCESS
> + fi
> +
> fi
>
> if [ "$slave_sql" != 'Yes' ]; then
> # We don't have a replication SQL thread running. Not a
> - # good thing. Try to recoved by restarting the resource in
> - # place.
> + # good thing. Try to recoved by restarting the SQL thread
> + # and remove reader vip. Prevent MySQL restart.
> ocf_log err "MySQL Slave SQL threads currently not running."
> ocf_log err "See $tmpfile for details"
> - exit $OCF_ERR_GENERIC
> +
> + # Remove reader vip
> + set_reader_attr 0
> +
> + # try to restart slave
> + ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
> + -e "START SLAVE"
> +
> + # Return success to prevent a restart
> + exit $OCF_SUCCESS
> fi
>
> if ocf_is_true $OCF_RESKEY_evict_outdated_slaves; then
> @@ -491,9 +539,13 @@
> if [ $secs_behind -gt $OCF_RESKEY_max_slave_lag ]; then
> ocf_log err "MySQL Slave is $secs_behind seconds behind master (allowed maximum: $OCF_RESKEY_max_slave_lag)."
> ocf_log err "See $tmpfile for details"
> +
> + # Remove reader vip
> + set_reader_attr 0
> +
> exit $OCF_ERR_INSTALLED
> fi
> - elif ocf_is_ms; then
> + elif ocf_is_ms; then
> # Even if we're not set to evict lagging slaves, we can
> # still use the seconds behind master value to set our
> # master preference.
> @@ -506,11 +558,19 @@
> $CRM_MASTER -v $master_pref
> fi
>
> + # is the slave ok to have a VIP on it
> + if [ $secs_behind -gt $OCF_RESKEY_max_slave_lag ]; then
> + set_reader_attr 0
> + else
> + set_reader_attr 1
> + fi
> +
> ocf_log debug "MySQL instance running as a replication slave"
> rm -f $tmpfile
> else
> # Instance produced an empty "SHOW SLAVE STATUS" output --
> # instance is not a slave
> + # TODO: Needs to handle when get_slave_info will return too many connections error
> rm -f $tmpfile
> ocf_log err "check_slave invoked on an instance that is not a replication slave."
> exit $OCF_ERR_GENERIC
> @@ -518,26 +578,26 @@
> }
>
> set_master() {
> - local new_master_host master_log_file master_log_pos
> + local new_master master_log_file master_log_pos
> local master_params
>
> - new_master_host=$1
> + new_master=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f1`
>
> # Keep replication position
> get_slave_info
>
> - if [ "$master_log_file" -a "$new_master_host" = "$master_host" ]; then
> -# master_params=", MASTER_LOG_FILE='$master_log_file', \
> -# MASTER_LOG_POS=$master_log_pos"
> + if [ "$master_log_file" -a "$new_master" = "$master_host" ]; then
> + # master_params=", MASTER_LOG_FILE='$master_log_file', \
> + # MASTER_LOG_POS=$master_log_pos"
> ocf_log info "Kept master pos for $master_host : $master_log_file:$master_log_pos"
> rm -f $tmpfile
> return
> else
> - master_log_file=`$CRM_ATTR -n $new_master_host-log-file-${INSTANCE_ATTR_NAME} -q -G`
> - master_log_pos=`$CRM_ATTR -n $new_master_host-log-pos-${INSTANCE_ATTR_NAME} -q -G`
> + master_log_file=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f2`
> + master_log_pos=`$CRM_ATTR_REPL_INFO --query -q | cut -d'|' -f3`
> if [ -n "$master_log_file" -a -n "$master_log_pos" ]; then
> master_params=", MASTER_LOG_FILE='$master_log_file', \
> - MASTER_LOG_POS=$master_log_pos"
> + MASTER_LOG_POS=$master_log_pos"
> ocf_log info "Restored master pos for $new_master_host : $master_log_file:$master_log_pos"
> fi
> fi
> @@ -548,11 +608,10 @@
> # from the laste master the slave replicated from, or freshly
> # reset with RESET MASTER.
>
> - ocf_run $MYSQL $MYSQL_OPTIONS_LOCAL $MYSQL_OPTIONS_REPL \
> - -e "CHANGE MASTER TO MASTER_HOST='$new_master_host', \
> - MASTER_USER='$OCF_RESKEY_replication_user', \
> - MASTER_PASSWORD='$OCF_RESKEY_replication_passwd' $master_params"
> -
> + ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
> + -e "CHANGE MASTER TO MASTER_HOST='$new_master', \
> + MASTER_USER='$OCF_RESKEY_replication_user', \
> + MASTER_PASSWORD='$OCF_RESKEY_replication_passwd' $master_params"
> rm -f $tmpfile
> }
>
> @@ -569,15 +628,37 @@
> return $OCF_SUCCESS
> fi
>
> - local mysql_options
> - mysql_options="$MYSQL_OPTIONS_LOCAL $MYSQL_OPTIONS_REPL"
> -
> local tmpfile
> tmpfile=`mktemp ${HA_RSCTMP}/unset_master.${OCF_RESOURCE_INSTANCE}.XXXXXX`
>
> - # First, stop the slave I/O thread and wait for relay log
> + # At this point, the master is read only so there should not be much binlogs to transfer
> + # Let's wait for the last bits
> + while true; do
> + $MYSQL $MYSQL_OPTIONS_REPL \
> + -e 'SHOW PROCESSLIST\G' > $tmpfile
> + if grep -i 'Waiting for master to send event' $tmpfile >/dev/null; then
> + ocf_log info "MySQL slave has finished reading master binary log"
> + break
> + fi
> + if grep -i 'Reconnecting after a failed master event read' $tmpfile >/dev/null; then
> + ocf_log info "Master is down, no more binary logs to come"
> + break
> + fi
> + if grep -i 'Connecting to master' $tmpfile >/dev/null; then
> + ocf_log info "Master is down, no more binary logs to come"
> + break
> + fi
> + if ! grep 'system user' $tmpfile >/dev/null; then
> + ocf_log info "Slave is not running - not waiting to finish"
> + break
> + fi
> +
> + sleep 1
> + done
> +
> + # Now, stop the slave I/O thread and wait for relay log
> # processing to complete
> - ocf_run $MYSQL $mysql_options \
> + ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
> -e "STOP SLAVE IO_THREAD"
> if [ $? -gt 0 ]; then
> ocf_log err "Error stopping slave IO thread"
> @@ -585,9 +666,9 @@
> fi
>
> while true; do
> - $MYSQL $mysql_options \
> + $MYSQL $MYSQL_OPTIONS_REPL \
> -e 'SHOW PROCESSLIST\G' > $tmpfile
> - if grep '[Hh]as read all relay log' $tmpfile >/dev/null; then
> + if grep -i 'Has read all relay log' $tmpfile >/dev/null; then
> ocf_log info "MySQL slave has finished processing relay log"
> break
> fi
> @@ -597,45 +678,93 @@
> fi
> ocf_log info "Waiting for MySQL slave to finish processing relay log"
> sleep 1
> - done
> + done
> rm -f $tmpfile
>
> # Now, stop all slave activity and unset the master host
> - ocf_run $MYSQL $mysql_options \
> + ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
> -e "STOP SLAVE"
> if [ $? -gt 0 ]; then
> ocf_log err "Error stopping rest slave threads"
> exit $OCF_ERR_GENERIC
> fi
>
> - #Save current state
> - get_slave_info
> - $CRM_ATTR -n $master_host-log-file-${INSTANCE_ATTR_NAME} -v $master_log_file
> - $CRM_ATTR -n $master_host-log-pos-${INSTANCE_ATTR_NAME} -v $master_log_pos
> - rm -f $tmpfile
> -
> - ocf_run $MYSQL $mysql_options \
> - -e "CHANGE MASTER TO MASTER_HOST=''"
> + ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
> + -e "RESET SLAVE;"
> if [ $? -gt 0 ]; then
> ocf_log err "Failed to set master"
> exit $OCF_ERR_GENERIC
> fi
> }
>
> -# Start replication as slave. Master hostname as parameter
> +# Start replication as slave
> start_slave() {
> - local master_host
> -
> - master_host="$1"
> -
> - # Remove state attributes - it will be invalid after START SLAVE
> - $CRM_ATTR -n $master_host-log-file-${INSTANCE_ATTR_NAME} -D
> - $CRM_ATTR -n $master_host-log-pos-${INSTANCE_ATTR_NAME} -D
>
> - ocf_run $MYSQL $MYSQL_OPTIONS_LOCAL $MYSQL_OPTIONS_REPL \
> + ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
> -e "START SLAVE"
> }
>
> +# Set the attribute controlling the readers VIP
> +set_reader_attr() {
> + local curr_attr_value
> +
> + curr_attr_value=$(get_reader_attr)
> +
> + if [ "$curr_attr_value" -ne "$1" ]; then
> + $CRM_ATTR -l reboot --name ${OCF_RESKEY_reader_attribute} -v $1
> + fi
> +
> +}
> +
> +# get the attribute controlling the readers VIP
> +get_reader_attr() {
> + local attr_value
> + local rc
> +
> + attr_value=`$CRM_ATTR -l reboot --name ${OCF_RESKEY_reader_attribute} --query -q`
> + rc=$?
> + if [ "$rc" -eq "0" ]; then
> + echo $attr_value
> + else
> + echo -1
> + fi
> +
> +}
> +
> +# Stores data for MASTER STATUS from MySQL
> +update_data_master_status() {
> +
> + master_status_file="${HA_RSCTMP}/master_status.${OCF_RESOURCE_INSTANCE}"
> +
> + $MYSQL $MYSQL_OPTIONS_REPL -e "SHOW MASTER STATUS\G" > $master_status_file
> +}
> +
> +
> +# Returns the specified value from the stored copy of SHOW MASTER STATUS.
> +# should be call after update_data_master_status for tmpfile
> +# Arguments:
> +# $1 The value to get.
> +get_master_status() {
> + awk -v var="$1" '$1 == var ":" {print substr($0, index($0, ":") + 2)}' "$master_status_file"
> +}
> +
> +
> +# Determines what IP address is attached to the current host. The output of the
> +# crm_attribute command looks like this:
> +# scope=nodes name=IP value=10.2.2.161
> +# If the ${INSTANCE_ATTR_NAME}_MYSQL_MASTER_IP node attribute is not defined, fallback is to uname -n
> +# The ${INSTANCE_ATTR_NAME}_MYSQL_MASTER_IP is the IP address that will be used for the
> +# change master to command.
> +get_local_ip() {
> + local IP
> + IP=`$CRM_ATTR -l forever -n ${INSTANCE_ATTR_NAME}_mysql_master_IP -q -G`
> + if [ ! $? -eq 0 ]; then
> + uname -n
> + else
> + echo $IP
> + fi
> +}
> +
> #######################################################################
>
> # Functions invoked by resource manager actions
> @@ -670,12 +799,8 @@
> }
>
> mysql_status() {
> - # Set the log level of the error message (default:err)
> - local loglevel
> - loglevel=${1:-err}
> -
> if [ ! -e $OCF_RESKEY_pid ]; then
> - ocf_log $loglevel "MySQL is not running"
> + ocf_log $1 "MySQL is not running"
> return $OCF_NOT_RUNNING;
> fi
>
> @@ -689,7 +814,7 @@
> if [ $? -eq 0 ]; then
> return $OCF_SUCCESS;
> else
> - ocf_log $loglevel "MySQL not running: removing old PID file"
> + ocf_log $1 "MySQL not running: removing old PID file"
> rm -f $OCF_RESKEY_pid
> return $OCF_NOT_RUNNING;
> fi
> @@ -703,10 +828,13 @@
> if ocf_is_probe; then
> status_loglevel="info"
> fi
> -
> +
> mysql_status $status_loglevel
> +
> rc=$?
>
> + # TODO: check max connections error
> +
> # If status returned an error, return that immediately
> if [ $rc -ne $OCF_SUCCESS ]; then
> return $rc
> @@ -720,16 +848,13 @@
> check_slave
> fi
>
> - local mysql_options
> - mysql_options="$MYSQL_OPTIONS_LOCAL --user=$OCF_RESKEY_test_user --password=$OCF_RESKEY_test_passwd"
> -
> # Check for test table
> - ocf_run -q $MYSQL $mysql_options \
> + ocf_run -q $MYSQL $MYSQL_OPTIONS_TEST \
> -e "SELECT COUNT(*) FROM $OCF_RESKEY_test_table"
> rc=$?
>
> if [ $rc -ne 0 ]; then
> - ocf_log err "Failed to select from $OCF_RESKEY_test_table";
> + ocf_log err "Failed to select from $test_table";
> return $OCF_ERR_GENERIC;
> fi
> fi
> @@ -744,6 +869,11 @@
> }
>
> mysql_start() {
> + if ocf_is_ms; then
> + # Initialize the ReaderVIP attribute, monitor will enable it
> + set_reader_attr 0
> + fi
> +
> mysql_status info
> if [ $? = $OCF_SUCCESS ]; then
> ocf_log info "MySQL already running"
> @@ -784,10 +914,10 @@
> # already existed, check whether it is writable by the configured
> # user
> for dir in $pid_dir $socket_dir; do
> - if ! su -s /bin/sh - $OCF_RESKEY_user -c "test -w $dir"; then
> - ocf_log err "Directory $dir is not writable by $OCF_RESKEY_user"
> - exit $OCF_ERR_PERM;
> - fi
> + if ! su -s /bin/sh - $OCF_RESKEY_user -c "test -w $dir"; then
> + ocf_log err "Directory $dir is not writable by $OCF_RESKEY_user"
> + exit $OCF_ERR_PERM;
> + fi
> done
>
> # Uncomment to perform permission clensing
> @@ -802,11 +932,11 @@
> fi
>
> ${OCF_RESKEY_binary} --defaults-file=$OCF_RESKEY_config \
> - --pid-file=$OCF_RESKEY_pid \
> - --socket=$OCF_RESKEY_socket \
> - --datadir=$OCF_RESKEY_datadir \
> - --user=$OCF_RESKEY_user $OCF_RESKEY_additional_parameters \
> - $mysql_extra_params >/dev/null 2>&1 &
> + --pid-file=$OCF_RESKEY_pid \
> + --socket=$OCF_RESKEY_socket \
> + --datadir=$OCF_RESKEY_datadir \
> + --user=$OCF_RESKEY_user $OCF_RESKEY_additional_parameters \
> + $mysql_extra_params >/dev/null 2>&1 &
> rc=$?
>
> if [ $rc != 0 ]; then
> @@ -823,8 +953,8 @@
> if [ $rc = $OCF_SUCCESS ]; then
> start_wait=0
>
> - elif [ $rc != $OCF_NOT_RUNNING ]; then
> - ocf_log err "MySQL start failed: $rc"
> + elif [ $rc != $OCF_NOT_RUNNING ]; then
> + ocf_log info "MySQL start failed: $rc"
> return $rc
> fi
> sleep 2
> @@ -836,30 +966,15 @@
> # already promoted a master. So, we simply start in read only
> # mode.
> set_read_only on
> -
> +
> # Now, let's see whether there is a master. We might be a new
> # node that is just joining the cluster, and the CRM may have
> # promoted a master before.
> master_host=`echo $OCF_RESKEY_CRM_meta_notify_master_uname|tr -d " "`
> if [ "$master_host" -a "$master_host" != ${HOSTNAME} ]; then
> ocf_log info "Changing MySQL configuration to replicate from $master_host."
> - set_master $master_host
> - ocf_run $MYSQL $MYSQL_OPTIONS_LOCAL $MYSQL_OPTIONS_REPL \
> - -e 'START SLAVE'
> - if [ $? -ne 0 ]; then
> - ocf_log err "Failed to start slave"
> - return $OCF_ERR_GENERIC
> - fi
> - else
> - ocf_log info "No MySQL master present, clearing replication state"
> - unset_master
> - fi
> -
> - master_host=`echo $OCF_RESKEY_CRM_meta_notify_master_uname`
> - if [ "$master_host" -a "$master_host" != ${HOSTNAME} ]; then
> - ocf_log info "Changing MySQL configuration to replicate from $master_host."
> - set_master $master_host
> - start_slave $master_host
> + set_master
> + start_slave
> if [ $? -ne 0 ]; then
> ocf_log err "Failed to start slave"
> return $OCF_ERR_GENERIC
> @@ -874,6 +989,7 @@
> # preference set by the administrator. We choose a low
> # greater-than-zero preference.
> $CRM_MASTER -v 1
> +
> fi
>
> # Initial monitor action
> @@ -886,7 +1002,7 @@
> ocf_log err "Failed initial monitor action"
> return $rc
> fi
> -
> +
> ocf_log info "MySQL started"
> return $OCF_SUCCESS
> }
> @@ -896,12 +1012,14 @@
> if ocf_is_ms; then
> # clear preference for becoming master
> $CRM_MASTER -D
> +
> + # Remove VIP capability
> + set_reader_attr 0
> fi
>
> - mysql_status info
> - rc=$?
> - if [ $rc = $OCF_NOT_RUNNING ]; then
> - return $OCF_SUCCESS
> + if [ ! -f $OCF_RESKEY_pid ]; then
> + ocf_log info "MySQL is not running"
> + return $OCF_SUCCESS
> fi
>
> pid=`cat $OCF_RESKEY_pid 2> /dev/null `
> @@ -911,14 +1029,14 @@
> ocf_log err "MySQL couldn't be stopped"
> return $OCF_ERR_GENERIC
> fi
> -
> # stop waiting
> shutdown_timeout=15
> if [ -n "$OCF_RESKEY_CRM_meta_timeout" ]; then
> shutdown_timeout=$((($OCF_RESKEY_CRM_meta_timeout/1000)-5))
> fi
> count=0
> - while [ $count -lt $shutdown_timeout ]; do
> + while [ $count -lt $shutdown_timeout ]
> + do
> mysql_status info
> rc=$?
> if [ $rc = $OCF_NOT_RUNNING ]; then
> @@ -926,12 +1044,12 @@
> fi
> count=`expr $count + 1`
> sleep 1
> - ocf_log info "MySQL still hasn't stopped yet. Waiting..."
> + ocf_log debug "MySQL still hasn't stopped yet. Waiting..."
> done
>
> mysql_status info
> if [ $? != $OCF_NOT_RUNNING ]; then
> - ocf_log warn "MySQL failed to stop after ${shutdown_timeout}s using SIGTERM. Trying SIGKILL..."
> + ocf_log info "MySQL failed to stop after ${shutdown_timeout}s using SIGTERM. Trying SIGKILL..."
> /bin/kill -KILL $pid > /dev/null
> fi
>
> @@ -942,11 +1060,20 @@
> }
>
> mysql_promote() {
> - if ( ! mysql_status ); then
> + local master_info
> +
> + if ( ! mysql_status err ); then
> return $OCF_NOT_RUNNING
> fi
> - ocf_run $MYSQL $MYSQL_OPTIONS_LOCAL $MYSQL_OPTIONS_REPL \
> + ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
> -e "STOP SLAVE"
> +
> + # Set Master Info in CIB, cluster level attribute
> + update_data_master_status
> + master_info="$(get_local_ip)|$(get_master_status File)|$(get_master_status Position)"
> + ${CRM_ATTR_REPL_INFO} -v "$master_info"
> + rm -f $tmpfile
> +
> set_read_only off || return $OCF_ERR_GENERIC
>
> # Existing master gets a higher-than-default master preference, so
> @@ -954,20 +1081,17 @@
> # unnecessarily
> $CRM_MASTER -v $((${OCF_RESKEY_max_slave_lag}+1))
>
> + # A master can accept reads
> + set_reader_attr 1
> +
> return $OCF_SUCCESS
> }
>
> mysql_demote() {
> - if ! mysql_status; then
> + if ! mysql_status err; then
> return $OCF_NOT_RUNNING
> fi
>
> - set_read_only on
> - if [ $? -ne 0 ]; then
> - ocf_log err "Failed to set read-only";
> - return $OCF_ERR_GENERIC;
> - fi
> -
> # Return master preference to default, so the cluster manager gets
> # a chance to select a new master
> $CRM_MASTER -v 1
> @@ -987,68 +1111,105 @@
> ocf_log debug "Received $type_op notification."
>
> case "$type_op" in
> - 'pre-promote')
> - # A new master is about to being promoted. It's not in
> - # read-write mode yet (that only occurs when it actually
> - # executes the promote action), so we can now safely
> - # connect to it and wait for it to start replicating.
> - local master_host
> - local master_status
> - master_host=`echo $OCF_RESKEY_CRM_meta_notify_promote_uname`
> -
> - if ( ! mysql_status ); then
> - return $OCF_NOT_RUNNING
> - fi
> - if [ -z "$master_host" ]; then
> - ocf_log err "Unable to determine master host!"
> - return $OCF_ERR_GENERIC
> - fi
> + 'pre-promote')
> + # Nothing to do now here, new replication info not yet published
>
> - if [ $master_host = ${HOSTNAME} ]; then
> - ocf_log info "This will be new master"
> - else
> - ocf_log info "Changing MySQL configuration to replicate from $master_host"
> - set_master $master_host
> - if [ $? -ne 0 ]; then
> - return $OCF_ERR_GENERIC
> + ;;
> + 'post-promote')
> + # The master has completed its promotion. Now is a good
> + # time to check whether our replication slave is working
> + # correctly.
> + master_host=`echo $OCF_RESKEY_CRM_meta_notify_promote_uname|tr -d " "`
> + if [ "$master_host" = ${HOSTNAME} ]; then
> + ocf_log info "This will be the new master, ignoring post-promote notification."
> else
> - return $OCF_SUCCESS
> + ocf_log info "Resetting replication"
> + unset_master
> + if [ $? -ne 0 ]; then
> + return $OCF_ERR_GENERIC
> + fi
> +
> + ocf_log info "Changing MySQL configuration to replicate from $master_host"
> + set_master
> + if [ $? -ne 0 ]; then
> + return $OCF_ERR_GENERIC
> + fi
> +
> + start_slave
> + if [ $? -ne 0 ]; then
> + ocf_log err "Failed to start slave"
> + return $OCF_ERR_GENERIC
> + fi
> fi
> - fi
> - ;;
> - 'post-promote')
> - # The master has completed its promotion. Now is a good
> - # time to check whether our replication slave is working
> - # correctly.
> - master_host=`echo $OCF_RESKEY_CRM_meta_notify_promote_uname`
> - if [ "$master_host" = ${HOSTNAME} ]; then
> - ocf_log info "Ignoring post-promote notification for my own promotion."
> return $OCF_SUCCESS
> - fi
> - start_slave $master_host
> - if [ $? -ne 0 ]; then
> - ocf_log err "Failed to start slave"
> - return $OCF_ERR_GENERIC
> - fi
> ;;
> - 'post-demote')
> - demote_host=`echo $OCF_RESKEY_CRM_meta_notify_demote_uname`
> - if [ $demote_host = ${HOSTNAME} ]; then
> - ocf_log info "Ignoring post-demote notification for my own demotion."
> + 'pre-demote')
> + demote_host=`echo $OCF_RESKEY_CRM_meta_notify_demote_uname|tr -d " "`
> + if [ $demote_host = ${HOSTNAME} ]; then
> + ocf_log info "post-demote notification for $demote_host"
> + set_read_only on
> + if [ $? -ne 0 ]; then
> + ocf_log err "Failed to set read-only";
> + return $OCF_ERR_GENERIC;
> + fi
> +
> + # Must kill all existing user threads because they are still Read/write
> + # in order for the slaves to complete the read of binlogs
> + local tmpfile
> + tmpfile=`mktemp ${HA_RSCTMP}/threads.${OCF_RESOURCE_INSTANCE}.XXXXXX`
> + $MYSQL $MYSQL_OPTIONS_REPL \
> + -e "SHOW PROCESSLIST" > $tmpfile
> +
> + for thread in `awk '$0 !~ /Binlog Dump|system user|event_scheduler|SHOW PROCESSLIST/ && $0 ~ /^[0-9]/ {print $1}' $tmpfile`
> + do
> + ocf_run $MYSQL $MYSQL_OPTIONS_REPL \
> + -e "KILL ${thread}"
> + done
> + else
> + ocf_log info "Ignoring post-demote notification execpt for my own demotion."
> + fi
> return $OCF_SUCCESS
> - fi
> - ocf_log info "post-demote notification for $demote_host."
> - # The former master has just been gracefully demoted.
> - unset_master
> ;;
> - *)
> - return $OCF_SUCCESS
> + 'post-demote')
> + demote_host=`echo $OCF_RESKEY_CRM_meta_notify_demote_uname|tr -d " "`
> + if [ $demote_host = ${HOSTNAME} ]; then
> + ocf_log info "Ignoring post-demote notification for my own demotion."
> + return $OCF_SUCCESS
> + fi
> + ocf_log info "post-demote notification for $demote_host."
> + # The former master has just been gracefully demoted.
> + unset_master
> + ;;
> + *)
> + return $OCF_SUCCESS
> ;;
> esac
> }
>
> #######################################################################
>
> +
> +##########################################################################
> +# If DEBUG_LOG is set, make this resource agent easy to debug: set up the
> +# debug log and direct all output to it. Otherwise, redirect to /dev/null.
> +# The log directory must be a directory owned by root, with permissions 0700,
> +# and the log must be writable and not a symlink.
> +##########################################################################
> +DEBUG_LOG="/tmp/mysql.ocf.ra.debug/log"
> +if [ "${DEBUG_LOG}" -a -w "${DEBUG_LOG}" -a ! -L "${DEBUG_LOG}" ]; then
> + DEBUG_LOG_DIR="${DEBUG_LOG%/*}"
> + if [ -d "${DEBUG_LOG_DIR}" ]; then
> + exec 9>>"$DEBUG_LOG"
> + exec 2>&9
> + date >&9
> + echo "$*" >&9
> + env | grep OCF_ | sort >&9
> + set -x
> + else
> + exec 9>/dev/null
> + fi
> +fi
> +
> case "$1" in
> meta-data) meta_data
> exit $OCF_SUCCESS;;
> @@ -1072,7 +1233,7 @@
> case "$1" in
> start) mysql_start;;
> stop) mysql_stop;;
> - status) mysql_status;;
> + status) mysql_status err;;
> monitor) mysql_monitor;;
> promote) mysql_promote;;
> demote) mysql_demote;;
> _______________________________________________________
> Linux-HA-Dev: Linux-HA-Dev [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
> Home Page: http://linux-ha.org/
_______________________________________________________
Linux-HA-Dev: Linux-HA-Dev [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
Home Page: http://linux-ha.org/
|
patch.ra.mysql
