7 | | * `mf-monitor-filtercheck` - This script runs on a cron job and updates the nagios server if the status changes. When run on the terminal it will take 15 minutes to complete (to leave enough time for all messages to be delivered) and will output either OK: or CRITICAL: depending on whether all messages laned in the inbox or not. |
8 | | * `mf-filtercheck` - This script is the one executed by `mf-monitor-filtercheck` - by default it provides verbose output so you can see which corporate providers are filtering our email. You can pass the environment variable `MF_SLEEP=60` if you want the script to sleep for only 60 seconds instead of the full 15 minutes. |
| 7 | * `mf-monitor-filtercheck` - This script runs on a cron job and updates the nagios server if the status changes. When run on the terminal it will take 45 minutes to complete (to leave enough time for all messages to be delivered) and will output either OK: or CRITICAL: depending on whether all messages laned in the inbox or not. |
| 8 | * `mf-filtercheck` - This script is the one executed by `mf-monitor-filtercheck` - by default it provides verbose output so you can see which corporate providers are filtering our email. You can pass the environment variable `MF_SLEEP=60` if you want the script to sleep for only 60 seconds instead of the full 45 minutes. |
| 10 | |
| 11 | === What to do if the nagios monitor alert goes off === |
| 12 | |
| 13 | Check the logs. |
| 14 | |
| 15 | {{{ |
| 16 | journalctl | grep filter-check |
| 17 | }}} |
| 18 | |
| 19 | Or |
| 20 | |
| 21 | {{{ |
| 22 | cat /var/log/syslog | grep filter-check |
| 23 | }}} |
| 24 | |
| 25 | If the exit code is "1" it means the message landed in the mailbox. If the exit code is 2 it means it was not received. If the exit code was 255 it means there was an error. |
| 26 | |
| 27 | If the exit code is "1" - you can retest to see if this is happening consistently: |
| 28 | |
| 29 | {{{ |
| 30 | filter-check --sendto SENDTO --sleep 30 |
| 31 | }}} |
| 32 | |
| 33 | If the exit code is "2" - you can manually login to the provider (see [wiki:keyringer] for all passwords) and see if the message eventually made it. Then you can debug the delay. |
| 34 | |
| 35 | If the exit code is "3" - you can try to re-run manually using `filter-check` (like with exit code 1) to see if you can re-create the error. |