Reset your server clocks when you get in

[shpilkus]

I have successfully SSH-ed into each of our 'virtualized' servers, and they all had different versions of the incorrect time. To reset the clock use the following at the SSH prompt (must be logged in as root):

Code: Select all

date -s "04 NOV 2007 17:45:00"

Adjust the date or time as required. And the quotes are needed.

[sixpackmx]

Thre should be a ntp server in place. I use time.nist.gov for syncing the clock.

[HMPGON]

What is the comand to just 'check' what the hell my server has as time & date without changing anyting?

[crashdump]

Code: Select all

date

[Greg]

i get a different format when i look at the date, so should i neter the new time/date should i enter the format i got from using "date" or should i use your format?

my current date is

Code: Select all

Tue Nov  6 13:22:22 EST 2007

So should i enter your format:

Code: Select all

date -s "04 NOV 2007 17:45:00"

where it is date-month-year-time


or enter in the same format as mine:

Code: Select all

date -s "Tue Nov  6 13:22:22 EST 2007"

where it is day of week-month-date-time-time zone-year

[Greg]

never mind, looked at the man pages, all i need to do was change the time, so i did:

date -s 17:33

[sixpackmx]

best way to do this:

ntpdate time.nist.gov

This will sync with the official NIST time.

[rldev]

best way to do this:

ntpdate time.nist.gov

This will sync with the official NIST time.

how do you always keep it synced? cron?

[shpilkus]

Supposed to by ntpd but the script to start/stop/restart it is broked. I can start it manually but on reboot it would be down (as is Apache).

Man, once everyone is restored they have a lot of cleanup and rolling out to do as a bunch of the automation is whacked.

[Sapphyre]

Supposed to by ntpd but the script to start/stop/restart it is broked. I can start it manually but on reboot it would be down (as is Apache).

Man, once everyone is restored they have a lot of cleanup and rolling out to do as a bunch of the automation is whacked.

Are you seeing this broken on a blade or a box ?

[shpilkus]

Blades.

[Sapphyre]

Blades.

So it wasn't as easy as copying all files including OS over. What happened with all that pretesting they did? No one noticed or thought of testing this stuff?

[Okidoki]

The cron job for root on physical server is delete o rename for root.bak

Crerate new cron for root and change this line

*/20 * * * * /usr/sbin/ntpdate -b 209.239.56.43 2>/dev/null >/dev/null

for example

*/20 * * * * /usr/sbin/ntpdate -b 192.43.244.18 2>/dev/null >/dev/null