It waits until a problem is real before it worries you.
PimiPulse is server monitoring that answers in sentences. The monitor reads your server every 10 seconds, and the verdict changes only when a problem has lasted.
Free for up to 3 servers. No card needed.
The monitor reads every 10 seconds and reports once a minute
It is one small program. It reads numbers Linux already keeps and sends one short report a minute.
Memory
How much is free, and whether the server is pushing memory out to disk.
Processor
How busy it is, and how much of it went to other customers on the same machine.
Every disk
How full each one is and how fast it is filling.
Programs and services
Which programs use the most, by name, and whether your web server, PHP, database and cache are running.
A busy minute is not a verdict
Those reports are read by 9 rules. Each one waits before it speaks, so a spike never becomes an alarm.
| What it sees | When it says so |
|---|---|
| Memory is tight | Amber when it stays tight. Red when it is nearly gone. |
| The processor is flat out | Amber when it stays busy, not for a spike. |
| Memory or the processor keeps peaking | Amber when short peaks keep coming back. |
| A backup is using the whole processor | Nothing, unless it runs long. |
| A web server or database stopped | Red when it stays down. Back within seconds: nothing. |
| The server ran out of memory | Red at once. |
| Reports stop arriving | Amber, then red if they stay away. Never red while your server still answers. |
Your server’s processor has been flat out for 20 minutes, mostly running your website (PHP).
Pages feel slow while this lasts. If it is a one-off, such as an import or a backup, it will pass on its own.
What to do: Nothing to do yet. If it happens most days, move to a plan with more processor power.
An email when it turns red, and another when it clears
Amber shows on your server’s page and in the Monday summary. It emails only for repeated peaks. Red emails at once, then again when it eases and when it is all clear.
Your server is struggling, here is what to do. Your server just ran out of memory and had to force-close your database to survive. Visitors probably saw errors for a moment. This will keep happening until the server has more memory, or until whatever grew is fixed. What to do: Upgrade memory from your hosting panel, or ask your developer what grew.
Here is how your server did this week (Mon 2 Nov to Mon 9 Nov). - my-shop: healthy all week. Checked 10,079 times; nothing needed you. Memory: usually about 745 MB free (38%); lowest 369 MB (19%) briefly on Fri 6 Nov at 00:00 UTC. Processor: usually 91% free; busiest 7% free briefly on Fri 6 Nov at 00:23 UTC. Disk: 27.2 GB left (61% free).
The red email goes on with a link to the server’s page and a short technical note for your developer.
A new server gets one email after its first 7 days: what we learned.
What it does not bother you with
- A nightly backup. A short backup raises nothing.
- A restart during updates. A database or web server that is back within seconds is not reported.
- A reboot after you resize. A server that comes back with more memory or more processors is read as an upgrade, not a crash.
- Fixing it for you. PimiPulse never changes anything on your server. It says what to do and stops there.
After 14 days, the server’s page says whether your size is right
It judges the busiest minute, not the usual day. The Monday summary says it at most every 4 weeks.
| What it finds | What it says |
|---|---|
| The busiest minute would not fit the next size down with room to spare | Right size, and when the smaller size would have fallen short |
| A database or cache among the biggest memory users | Right size: a database or cache grows to fill the memory it has |
| A program stopped because memory ran out, swap use that grew, or missed readings | Right size: a smaller size is not a safe call |
| Room to spare on the smaller size, and none of the above | A smaller size would likely do |
| A memory or processor warning is open | Nothing. The warning already says what to do. |
2 GB is the right size for this server.
At its busiest in the last 14 days this server used 1.4 GB of its 2 GB memory and about 1 of its 2 processors. The next size down (1 GB) would have run out on Tue 3 Nov at 14:12.
It tells you whether the upgrade worked
Mark a server as upgraded, or let PimiPulse notice the bigger size by itself. It puts the old warning aside, compares a full day on the new size with the day before, and answers on the server’s page.
Your upgrade worked.
Before the upgrade, memory was getting tight. In the day since the upgrade, that has not come back.