Using sysfs for monitoring of Linux Soft RAID aka MD
Tested on Zabbix Version 3.0
Automatically converted for Zabbix v2.0 and v2.4
See at share.zabbix.com
- LLD for auto detecting RAID
- Monitoring of total number of disks, number of degraded disks, sync_status
- Graph for degraded disks
- Triggers for degrading and raid recovering
- SYSFS at agents's server, /sys/block/md* available
- Agent (host with RAID)
- Copy userparameter_md.conf to /etc/zabbix/zabbix_agentd.d/userparameter_md.conf
curl -Ls https://git.io/fN9H5 | sudo tee /etc/zabbix/zabbix_agentd.d/userparameter_md.conf
- Restart zabbix_agent
- UBUNTU/DEBIAN:
sudo service zabbix-agent restart
- UBUNTU/DEBIAN:
- Copy userparameter_md.conf to /etc/zabbix/zabbix_agentd.d/userparameter_md.conf
- Server
- Import template template_md_x.x.xml file via web admin (x.x this is your version of Zabbix)
- Save the XML template file to a convenient location (or this archive)
- Open the Zabbix console
- Navigate to Configuration > Templates
- Click the Import button
- Select the downloaded template_md_x.x.xml file
- Click Import
- Import template template_md_x.x.xml file via web admin (x.x this is your version of Zabbix)
- md.discover - LLD data (MD RAIDs and disks)
- md.degraded[*] - number of degraded disks for specific RAID
- md.sync_action[*] - current sync status for specific RAID
- md.raid_disks[*] - number of all disks for specific RAID
- Add new UserParameter to /etc/zabbix/zabbix_agentd.d/userparameter_md.conf
- For example
UserParameter=md.sync_speed[*],cat /sys/block/$1/md/sync_speed
- Create new Item by instruction
Roman Kudlay roman@kudlay.pro
GNU GENERAL PUBLIC LICENSE Version 3, 29 June 2007
See LICENSE