diff --git a/xCAT-rmc/plugin/rmcmon.pm b/xCAT-rmc/plugin/rmcmon.pm index acaf5e85c..e11973b8c 100644 --- a/xCAT-rmc/plugin/rmcmon.pm +++ b/xCAT-rmc/plugin/rmcmon.pm @@ -30,7 +30,7 @@ use xCAT::DBobjUtils; #------------------------------------------------------------------------------- -=head1 xCAT_monitoring:rmcmon +=head1 xCAT_monitoring:rmcmon =head2 Package Description xCAT monitoring plugin package to handle RMC monitoring. =cut @@ -41,7 +41,7 @@ use xCAT::DBobjUtils; =head3 start This function gets called by the monitorctrl module - when xcatd starts and when monstart command is issued by the user. + when xcatd starts and when monstart command is issued by the user. It starts the daemons and does necessary startup process for the RMC monitoring. It also queries the RMC for its currently monitored nodes which will, in tern, compared with the nodes @@ -51,12 +51,12 @@ use xCAT::DBobjUtils; in the xCAT cluster. p_nodes -- a pointer to an arrays of nodes to be monitored. null means all. scope -- the action scope, it indicates the node type the action will take place. - 0 means localhost only. - 2 means both localhost and nodes, + 0 means localhost only. + 2 means both localhost and nodes, callback -- the callback pointer for error and status displaying. It can be null. Returns: - (return code, message) - if the callback is set, use callback to display the status and error. + (return code, message) + if the callback is set, use callback to display the status and error. =cut #-------------------------------------------------------------------------------- @@ -225,17 +225,17 @@ sub start { #-------------------------------------------------------------------------------- =head3 stop - This function gets called when monstop command is issued by the user. + This function gets called when monstop command is issued by the user. It stops the monitoring on all nodes, stops the RMC daemons. Arguments: p_nodes -- a pointer to an arrays of nodes to be stopped for monitoring. null means all. scope -- the action scope, it indicates the node type the action will take place. - 0 means localhost only. - 2 means both monservers and nodes, + 0 means localhost only. + 2 means both monservers and nodes, callback -- the callback pointer for error and status displaying. It can be null. Returns: - (return code, message) - if the callback is set, use callback to display the status and error. + (return code, message) + if the callback is set, use callback to display the status and error. =cut #-------------------------------------------------------------------------------- @@ -350,15 +350,15 @@ sub stop { #-------------------------------------------------------------------------------- =head3 config - This function configures the cluster for the given nodes. + This function configures the cluster for the given nodes. This function is called when moncfg command is issued or when xcatd starts on the service node. It will configure the cluster to include the given nodes within - the monitoring doamin. + the monitoring doamin. Arguments: p_nodes -- a pointer to an arrays of nodes to be added for monitoring. none means all. scope -- the action scope, it indicates the node type the action will take place. - 0 means localhost only. - 2 means localhost and nodes, + 0 means localhost only. + 2 means localhost and nodes, callback -- the callback pointer for error and status displaying. It can be null. Returns: (error code, error message) @@ -485,14 +485,14 @@ sub config { #-------------------------------------------------------------------------------- =head3 deconfig - This function de-configures the cluster for the given nodes. - This function is called when mondecfg command is issued by the user. + This function de-configures the cluster for the given nodes. + This function is called when mondecfg command is issued by the user. It should remove the given nodes from the product for monitoring. Arguments: p_nodes -- a pointer to an arrays of nodes to be removed for monitoring. none means all. scope -- the action scope, it indicates the node type the action will take place. - 0 means local host only. - 2 means both local host and nodes, + 0 means local host only. + 2 means both local host and nodes, callback -- the callback pointer for error and status displaying. It can be null. Returns: (error code, error message) @@ -563,11 +563,11 @@ sub deconfig { =head3 supportNodeStatusMon This function is called by the monitorctrl module to check if RMC can help monitoring and returning the node status. - + Arguments: none Returns: - 1 + 1 =cut #-------------------------------------------------------------------------------- @@ -584,14 +584,14 @@ sub supportNodeStatusMon { =head3 startNodeStatusMon This function is called by the monitorctrl module to tell RMC to start monitoring the node status and feed them back - to xCAT. RMC will start setting up the condition/response - to monitor the node status changes. + to xCAT. RMC will start setting up the condition/response + to monitor the node status changes. Arguments: p_nodes -- a pointer to an arrays of nodes for monitoring. none means all. scope -- the action scope, it indicates the node type the action will take place. - 0 means local host only. - 2 means both local host and nodes, + 0 means local host only. + 2 means both local host and nodes, callback -- the callback pointer for error and status displaying. It can be null. Returns: (error code, error message) @@ -708,7 +708,7 @@ sub startNodeStatusMon { Arguments: $oldstatus a pointer to a hash table that has the current node status - $node the name of the service node to run RMC command from. If null, get from local host. + $node the name of the service node to run RMC command from. If null, get from local host. Returns: (return code, message) =cut @@ -791,8 +791,8 @@ sub saveRMCNodeStatusToxCAT { Arguments: p_nodes -- a pointer to an arrays of nodes for monitoring. none means all. scope -- the action scope, it indicates the node type the action will take place. - 0 means local host only. - 2 means both local host and nodes, + 0 means local host only. + 2 means both local host and nodes, callback -- the callback pointer for error and status displaying. It can be null. Returns: (return code, message) @@ -924,10 +924,10 @@ sub getLocalNodeID { #-------------------------------------------------------------------------------- =head3 getNodeInfo - This function gets the nodeid, node ip addresses for the given node + This function gets the nodeid, node ip addresses for the given node Arguments: node - flag --- if 0 means normal nodes, if 1 means HMC. + flag --- if 0 means normal nodes, if 1 means HMC. Returns: (nodeid, nodeip) if error, nodeid=-1 and nodeip is the error message =cut @@ -972,10 +972,10 @@ sub getNodeInfo #-------------------------------------------------------------------------------- =head3 reportError - This function writes the error message to the callback, otherwise to syslog. + This function writes the error message to the callback, otherwise to syslog. Arguments: error - callback + callback Returns: none =cut @@ -997,13 +997,13 @@ sub reportError =head3 addNodes This function adds the nodes into the RMC cluster, it does not check the OSI type and - if the node has already defined. + if the node has already defined. Arguments: - nodes --an array of nodes to be added. + nodes --an array of nodes to be added. master -- the monitoring master of the node. scope -- the action scope, it indicates the node type the action will take place. - 0 means localhost only. - 2 means localhost and nodes, + 0 means localhost only. + 2 means localhost and nodes, callback -- the callback pointer for error and status displaying. It can be null. flag -- 0 means normal node. 1 means HMC. Returns: @@ -1243,11 +1243,11 @@ sub addNodes { =head3 removeNodes This function removes the nodes from the RMC cluster. Arguments: - nodes --a pointer to a array of node names to be removed. + nodes --a pointer to a array of node names to be removed. master -- the master of the nodes. scope -- the action scope, it indicates the node type the action will take place. - 0 means localhost only. - 2 means localhost and nodes, + 0 means localhost only. + 2 means localhost and nodes, callback -- the callback pointer for error and status displaying. It can be null. flag -- 0 means normal node. 1 means HMC. Returns: @@ -1410,7 +1410,7 @@ sub removeNodes { #-------------------------------------------------------------------------------- =head3 processSettingChanges - This function gets called when the setting for this monitoring plugin + This function gets called when the setting for this monitoring plugin has been changed in the monsetting table. Arguments: none. @@ -1427,7 +1427,7 @@ sub processSettingChanges { =head3 getDiscription This function returns the detailed description of the plugin inluding the - valid values for its settings in the monsetting tabel. + valid values for its settings in the monsetting tabel. Arguments: none Returns: @@ -1438,31 +1438,31 @@ sub processSettingChanges { sub getDescription { return " Description: - rmcmon uses IBM's Resource Monitoring and Control (RMC) component - of Reliable Scalable Cluster Technology (RSCT) to monitor the - xCAT cluster. RMC has built-in resources such as CPU, memory, - process, network, file system etc for monitoring. RMC can also be - used to provide node liveness status monitoring and performance monitoring for xCAT. - RMC is good for threadhold monitoring. xCAT automatically sets up the + rmcmon uses IBM's Resource Monitoring and Control (RMC) component + of Reliable Scalable Cluster Technology (RSCT) to monitor the + xCAT cluster. RMC has built-in resources such as CPU, memory, + process, network, file system etc for monitoring. RMC can also be + used to provide node liveness status monitoring and performance monitoring for xCAT. + RMC is good for threadhold monitoring. xCAT automatically sets up the monitoring domain for RMC during node deployment time. And if performance monitoring is enabled, xCAT monitoring will collect and consolidate performance - data based on montype&rmetrics setting, and then store to RRD database. + data based on montype&rmetrics setting, and then store to RRD database. Settings: - rfanout -- indicating the fanout number for configuring or deconfiguring + rfanout -- indicating the fanout number for configuring or deconfiguring remote nodes. nodeusebatch -- use RMC event batching function when doing node status monitoring. - montype -- indicating type of performance monitoring, the value could be + montype -- indicating type of performance monitoring, the value could be \"event\" : event monitoring (default), \"perf\" : performance monitoring, \"event,perf\" or \"perf,event\" : both. - rmetrics -- defines resource class,resouce names, attributes and sample interval for - performance monitoring. + rmetrics -- defines resource class,resouce names, attributes and sample interval for + performance monitoring. Syntax of key: rmetrics_resource class For example: rmetrics_IBM.Processor Syntax of value: [resource names]attribute names:sample interval The unit of sample interval is minute. For example: [proc0,proc1]PctTimeIdle,PctTimeWait,PctTimeKernel,PctTimeUser:5 - + "; } @@ -1470,10 +1470,10 @@ sub getDescription { =head3 getNodeConfData This function gets a list of configuration data that is needed by setting up - node monitoring. These data-value pairs will be used as environmental variables + node monitoring. These data-value pairs will be used as environmental variables on the given node. Arguments: - pointet to a arry of nodes + pointet to a arry of nodes Returns: pointer to a 2-level hash. For format will as following: { @@ -1529,7 +1529,7 @@ sub getNodeConfData { =head3 getPostscripts This function returns the postscripts needed for the nodes and for the servicd - nodes. + nodes. Arguments: none Returns: @@ -1735,15 +1735,15 @@ sub showeventlog #-------------------------------------------------------------------------------- =head3 show - This function configures the cluster for the given nodes. + This function configures the cluster for the given nodes. This function is called when moncfg command is issued or when xcatd starts on the service node. It will configure the cluster to include the given nodes within - the monitoring doamin. + the monitoring doamin. Arguments: p_nodes -- a pointer to an arrays of nodes to be added for monitoring. none means all. scope -- the action scope, it indicates the node type the action will take place. - 0 means localhost only. - 2 means localhost and nodes, + 0 means localhost only. + 2 means localhost and nodes, callback -- the callback pointer for error and status displaying. It can be null. Returns: (error code, error message)