Working out DownTime Averages..

[davieboy]

If i have a cron job set to ping a server / ip address and it is down, i want to be able to record to see how long it is down for and work out the average percentage relating to how long it has been monitored.

How do i go about doing this?

[slak]

Simple math?

1.Record when you first started monitoring.

2.Record the date/time when the site went down.
3.Record the date/time when it came back up.
4.Do your calculation.
5. Go to 2.

What if the server doesn't accept ICMP packets? It'll have 100% downtime.

[davieboy]

i can ping them fine just now, i just wanna know how to record the data.

How would i do the calculation

(never was good at math)

[slak]

From the time you started recording uptime, count minutes or seconds to the current time.

From the time the site went down, count the same minutes or seconds until it goes back online.

You'll have total uptime to date (a) and downtime (b)

Simply work out the percentage (b) is from (a).

Remember that (a) should continually be updated to give a real figure for (b).

[davieboy]

i take it is is best to record uptime in seconds? ten divide by 60 and then 24?

I really dont have a clue on this thing

lol

[daveyboy]

you should record it in seconds cos that is most compatible with the way the c library and php store time. you can always change it to hours, mins etc. to display, but the maths is easier to do with just seconds.

[davieboy]

ok kool

is this easy to do?

ps : did i nick the spelling of the name u wanted
😃

[daveyboy]

seconds is the easiest way: calling time() in php returns the time in seconds. for unix systems, the time is represented as the number of seconds since 1 Jan 1970. doing sums with hours, mins and seconds is hard.

re: the nickname. err no, i joined the forum over a year earlier than you.

[davieboy]

lol

😃

need to work out how to start it, got a rough idea