You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
# /etc/crontab: system-wide crontab
# Unlike any other crontab you don't have to run the `crontab'
# command to install the new version when you edit this file
# and files in /etc/cron.d. These files also have username fields,
# that none of the other crontabs do.
SHELL=/bin/sh
PATH=/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin
# m h dom mon dow user command
17 * * * * root cd / && run-parts --report /etc/cron.hourly
25 8 * * * root test -x /usr/sbin/anacron || ( cd / && run-parts --report /etc/cron.daily )
47 8 * * 7 root test -x /usr/sbin/anacron || ( cd / && run-parts --report /etc/cron.weekly )
52 8 1 * * root test -x /usr/sbin/anacron || ( cd / && run-parts --report /etc/cron.monthly )
The hour in this example isn't randomized. If you install 50 virtual machines on one hypervisor, all VMs will run their crons at the same time. This will lead into IO issues. we need to:
Check which operating systems are affected by this (probably every that ships a cron daemon)
create a function to randomize the time for each installation between 1-24 o'clock.
We've already own a Debian specific function for randomizing the mdadm check time:
The text was updated successfully, but these errors were encountered:
I think there are some implications here that users could depend on.
E.g.:
In all cases hourly jobs run before daily jobs, those run before weekly jobs and so on.
I also think that it is expected that those jobs run relatively early in the day, so the hour should be in the range of about 5 to 8.
For the minutes I would suggest to start with a random value from 0 to 19 for the minute of the hourly jobs, than add random 5 to 10 for daily, add random 5 to 10 for weekly, add random 5 to 10 for monthly.
I'm not sure about changing the minute at all because this adds a higher mount of complexity. what about running the jobs in the evening? between 10pm-8am for example? This should offer a range that is wide enough.
An example Debian
/etc/crontab
:The hour in this example isn't randomized. If you install 50 virtual machines on one hypervisor, all VMs will run their crons at the same time. This will lead into IO issues. we need to:
We've already own a Debian specific function for randomizing the mdadm check time:
The text was updated successfully, but these errors were encountered: