Monitoring
Monitoring & Alerts
Metrics for every server and database, with alert rules sent to Telegram or email.
What monitoring does
Monitoring shows how your servers are doing and tells you in Telegram or by email when something needs attention. It lives in the Cloud Panel under Monitoring.
The data has two sources. The hypervisor sees your server from outside and reports CPU, disk I/O and network every few seconds. The guest agent reports from inside once a minute, down to the space of each mount and the latency of each disk.
Both are on by default for servers from our OS templates. A server installed from your own ISO has hypervisor metrics only.
Why ours
Included with every server
Monitoring costs nothing extra. On servers from our OS templates it is on from the start.
Rules that stay quiet
We made our own set of rules that fits most servers. They catch real trouble and stay quiet on normal load, and the first server of an account gets them by itself.
One message per change
An alert sends one message when the problem starts, one more if it goes from warning to critical, and one when it is resolved. It does not repeat while the problem lasts.
Rules by tag
A rule covers all servers, servers with a tag like env=prod, or selected servers. Any rule can leave out servers with some tags.
How an alert behaves
An alert resolves when the reading drops 5 points below the warning threshold, so a server hovering at the threshold does not send a message every minute. If the same alert opens again within 30 minutes, it continues quietly.
When several rules fire on one server, you get one alert that names all of them, with the severity of the strictest. The message carries the server name, its public IP, the reading and the thresholds, with a link to the server.
A Telegram message has buttons to silence the alert for an hour or a day. A silenced alert is still recorded and shown on the Alerts tab, it only sends nothing.
A disk that grows
A full root disk takes a server down in odd ways: databases stop writing and packages fail to install. Auto-extend watches how full / is and, at the threshold you set, grows the disk together with the partition and the filesystem inside. The server keeps running, there is no reboot.
You set the threshold, the step and the largest size the disk may reach. After a failure auto-extend does not retry. It stops, tells you why and waits for you.
It works on servers from our OS templates with Monitoring on. Only the root disk grows.
Also in monitoring
All servers on one page
Charts across all monitored servers, and one row per server with its CPU, memory, disk space and agent state.
Alert history
Active alerts on top, then history, with search and filters. Silenced alerts are there too.
Pressure and OOM kills
Memory and I/O pressure from inside the server, and an alert when the kernel kills a process for lack of memory.
CPU Guard
On burst servers you hear when CPU Guard held the server at its guaranteed share for 10 minutes or more.
Alerts never go quiet
When the last working Telegram leaves the account, rules that sent to Telegram switch to the owner’s email until it is back.
Average over a time
A threshold rule averages the reading over 1 to 30 minutes before it compares it with the levels.
Questions
Does it cost anything?
No. Monitoring is included with every server.
Do I need to install an agent?
Not on servers from our OS templates, the agent and Monitoring are on by default. A server installed from your own ISO has hypervisor metrics only.
Who gets the alerts?
Telegram messages go to everyone in the account who linked @serverscamp_bot with Servers notifications on. Email goes to the account owner.
Can I mute an alert?
Yes. Mute it for an hour or a day from the Telegram message, or set a silence in Alert settings. It still shows on the Alerts tab.
Does auto-extend grow extra volumes?
No. Only the root disk grows.