--- name: "moncmd" section: "1" platform: "linux" package: "mon" version: "1.4.1-2" release: "Debian 13 (Trixie)" platforms: ["linux"] url: "https://new.carta.tech/moncmd" --- # moncmd(1) send commands to mon daemon and show the results. ## NAME moncmd - send commands to mon daemon and show the results. ## SYNOPSIS **moncmd** \[**-a**\] \[**-d**\] \[**-l** *username*\] \[**-f** *file*\] \[**-s** *server*\] \[**-p** *port*\] \[**command**\] ## DESCRIPTION **moncmd** sends commands to the **mon** server. ## OPTIONS - **-a** Authenticate with the server. - **-d** enable debugging, which is the same as **-s** *localhost*. - **-f** *file* Read and execute commands from *file*. - **-l** *username* Supply *username* as the username while authenticating to the server. - **-s** *server* Connect to *server*. - **-p** *port* Use TCP port *port* when connecting to the server, instead of the default of 32777. ## MONITOR HOST **moncmd** will use the host specified by the **-s** parameter as the server. If there is no **-s** parameter it will use the host specified in the MONHOST environment variable. If there is no host in either of these locations it will exit with an error. ## BATCH OPERATION If no commands are supplied to **moncmd** on the command line, then commands will be taken from either standard input, or from the file specified by the -f parameter. If standard input is connected to a TTY and the -a option is supplied, then it will prompt for a password. If the -a option is supplied without the -f option and standard input is not a TTY, then the username and password are read from standard input using the syntax "USER=username" and "PASS=password". The remaining input lines are interpreted as commands to send to the server. **moncmd** will not take usernames or passwords from a file, for obvious security reasons. If the username is neither supplied by the -l parameter or through standard input, it is taken from the effective user ID of the current process. ## COMMANDS The following is a list of the commands that the server understands. - **enable \| disable service** *group* *servicename* Enables/disables alerts for *group* and *servicename*. All disabled states are automatically saved to the state file, which may optionally be re-loaded upon restarting or initial startup of the server. - **ack** *group service comment* Acknowledge a failure condition. This will store *comment* in the state of service (queryable by doing a *list opstatus* command), and will surppress further alerts for the service. Once the service returns to a non-failure state, then the acknowledgement is reset. - **version** Displays the protocol version in the form of "version **num**" where **num** is the protocol version number. - **list aliases** Lists aliases. - **list aliasgroups** Lists alias groups. - **savestate** Save the state of the server. Currently, the only state which is saved is the host/watches/services which are disabled. - **loadstate** Load the state of the server. Currently, the only state which is loaded is the host/watches/services which are disabled. - **enable \| disable host** *hostname* Enables/disables host *hostname* in all groups. When the monitor is called, this hostname will not be included in the list of hostnames sent to the monitor. If a group has only one hostname in it, then the **enable \| disable watch** command should be used instead. - **enable \| disable watch** *watchgroup* Enables/disables an entire watch for *watchgroup*, as defined in the configuration file. Disabling a watch not only stops alerts from happening, but it stops the actual monitor processes from being scheduled. - **reset** Resets the server, forcing it to re-read the configuration file, kill off currently running monitors, restart all monitoring, and reset all counters. This command is only accessible if **moncmd** connects from the host which is running the **mon** server. - **reset stopped** Resets the server and immediately stops the scheduler. This is an atomic version of the commands **reset** and **stop**. - **reload auth** Reloads the auth.cf file in order to incorporate any new changes. The auth table is completely re-generated; it is not merged. - **reset keepstate** If the word "keepstate" comes after the reset command, the server will do a "loadstate" right after the reset, before the scheduler kicks back in. - **stop** Stops the scheduler, but continues to allow client connections. - **start** Re-starts the scheduler after it has been stopped. - **test monitor** *group* *servicename* Triggers a test for *group* and *service* immediately by setting the service's countdown timer to zero. - **test (alert \| upalert \| startupalert)** *group* *servicename* *retval* *period* Triggers a test alert, upalert, or startupalert for *group* and *servicename*. *retval* is the integer exit value to pass to the alert via the MON\_RETVAL environment variable. You must also specify the *period* as it appears in the configuration file. All alerts of the given type in that period will be triggered, but the alert will not be logged. - **servertime** Returns the current time of the server as seconds since Jan 1, 1970. - **list group** *groupname* Lists the members of group *groupname*. - **list descriptions** List the descriptions of each service, as defined in the configuration file. If a service description is undefined, then it is not listed. - **list alerthist** Lists the last alarms triggered for each service of each watch group, in addition to the summary output. The number of alerts to keep in memory is bounded by the *maxkeep* variable, configurable on the **mon** command line at startup, and expandable with the **set** command during runtime. - **list failurehist** Lists the last failures, in addition to the summary output. This is also limited by the *maxkeep* variable. - **list opstatus** Lists operational status of all services. Reports whether the last time a service group was tested resulted in success or failure. The output is: ``` group service untested group service time timeleft succeeded group service time timeleft failed output ``` where *output* is the first line of output from the monitor script which failed, *time* is the time that the condition was last noticed in [**time**(2)](/time/2.md) format, and *timeleft* is the number of seconds left until the service is tested again. - **list successes** Generates the same output as the **list opstatus** command, but only shows the services that have succeeded the last time they were tested. - **list failures** Generates the same output as the **list opstatus** command, but only shows the services that have failed the last time they were tested. - **list disabled** Lists all hosts and services which have been disabled by the **disable host\|service** command. - **list pids** Shows the currently active watch groups/services along with their process IDs, and it process ID of the server daemon. - **list watch** Lists all watches and services. - **list state** Lists the state of the scheduler. - **set** *group* *service* *variable* *value* Sets a variable to value. Useful for temporarily changing an interval or alertevery value. Be careful, because this can just set any value in the %watch hash, and some values that are specified in the configuration file like "10m" or "35s" are converted and stored as just plain integer seconds (e.g. "alertevery"). - **get** *group* *service* *variable* Displays the value of group service variable. - **set** *variable* *value* Assigns *value* to the global variable *variable*. - **set opstatus** *group* *service* *value* Sets the opstatus value for *group* and *service.* - **get** *variable* Shows the value of global variable *variable*. - **term** Terminates the server. This command is only accessible if **moncmd** connects from the host which is running the **mon** server. ## ENVIRONMENT VARIABLES - MONSERVER The hostname of the server which runs the **mon** process. - MONPORT The port number to connect to. ## SEE ALSO [mon(8)](/mon/8.md) ## BUGS Report bugs to the email address below. ## AUTHOR Jim Trocki \