Easy monitoring for YunoHost
This script allow you to monitor easily your servers. You can check outside the server to monitor:
And inside the server to monitor:
If you are using YunoHost, all this things are configured automatically from the configuration of the YunoHost to monitor.
This script supports ipv4 and ipv6.
It is extensible to add other kind of checks plugin and other kinds of actions (alert) plugin.
On the server to monitor:
*/8 * * * * some_user python3 yunomonitor.py -e MONITORING_SERVER [...]
On the monitoring server:
*/8 * * * * some_user python3 yunomonitor.py [SERVER_TO_MONITOR ...] [-m MAIL [MAIL ...]] [-s URL_SMS] [-c URL_CACHET]
This script is designed to run with these dependencies:
If you want to use the autoconfiguration mechanism, you need to configure on nginx a HTTPS .well-known/yunomonitor/
path. You need also in this case to have on monitoring server a ssh key on /etc/ssh/ssh_host_rsa_key
Here is a list of all checks methods. Each method return a list of small messages that describe errors append.
NO_IPV4_PING: NO_IPV6_PING:
C_DOMAIN_UNCONFIGURED C_DOMAIN_UNCONFIGURED_IN_IPV4 I_DOMAIN_UNCONFIGURED_IN_IPV6
E_CERT_RENEWED_FAILED C_CERT_INVALID C_PORT_CLOSED_OR_SERVICE_DOWN
C_TIMEOUT, ip, msg: C_TOO_MANY_REDIRECTS, ip, msg: C_SSO_CAPTURE: C_UNKNOWN_ERROR, ip, msg: C_HTTP_XXX, msg:
I_DOMAIN_EXPIRATION_NOT_FOUND W_DOMAIN_WILL_EXPIRE E_DOMAIN_NEARLY_EXPIRE C_DOMAIN_EXPIRE
C_BROKEN_NAMESERVER C_TIMEOUT C_DOMAIN_UNCONFIGURED C_NO_ANSWER C_UNEXPECTED_ANSWER
E_NO_MX_RECORD C_DOMAIN_UNCONFIGURED C_REVERSE_MISSING C_REVERSE_MISMATCH C_BLACKLISTED E_CERT_RENEWED_FAILED C_PORT_CLOSED_OR_SERVICE_DOWN
C_NOT_FOUND C_DOWN C_FAILED
W_SMART_NOT_SUPPORTED E_SMART_DISABLED E_SMART_HALF_WORKING C_DISK_FAILURE
C_FREE_SPACE E_FREE_SPACE W_FREE_SPACE
E_NEVER_BACKUP E_MISSING_BACKUP E_BACKUP_NOT_TRIGGERED E_BACKUP_BROKEN
W_APP_NEED_UPGRADE W_PKG_NEED_UPGRADE
This module send one email by failures checks.
This module send a condensate of all failures on all servers in one sms (sms apis are rarely free).
This module update components in cachet and open incident with information that could be shown publicly.
Because in some case, I need to deploy it quickly. An unique file is simple.
I have analyzed solution like nagios and zabbix, and these tools are very powerfull, but the cost to deploy and configure it is to high for my objective to deploy YunoHost with monitoring and backup automatically.
You can create a configuration file that will be used instead of the one generated by placing it in /etc/yunomonitor/DOMAIN.conf
ping:
- some.domain.tld
- other.domain.tld
https_200:
- wikipedia.org
- your.website.org
dns_resolver:
# Your favorite DNS resolver which you want to check is up
- 11.22.33.44
- 66.77.88.99
free_dns_service:
# Will test that a specific resolver correctly resolves a specific domain
- [ 'your.resolver.org', 'some.somain.tld', '12.34.56.78' ]
disk_health: True
free_space:
warning:
danger:
__components__:
- id: nextcloud__2
name: Nextcloud
label: Projet Vert
uris:
- other.domain.tld/
services:
- nginx
- id: borg
name: BorgBackup
label: Sauvegarde 1
backup: all
uris:
services: