Recently while working on a Project which had a hosting on Amazon Cloudfront, Storage on S3 and Streaming from Amazon EC2.I came across an issue which went ok initially but became a 2 day R&D Topic for my team after a server shutdown and Reset...
While requesting bucket I got a strange error from Amazon
"RequestTimeTooSkewed The difference between the request time and the current time is too large"
I checked everything from security credentials to bucket name finally going to Amazon Support where I came to knew that i was the fault of my system clock.
Our Systems must sync to NTP-Network Time Protocol in order to sync with Amazon Cloud Services..
My server
Centos which is redhat/suse based
Here is how i resolved the same
First and foremost I installed ntp again to be on safer side and make sure every stone in path was carved to needs
Install NTP
Yum install ntp
Check from config NTP is On
Chkconfig ntpd on
If everything is going well without any warnings till now its time to finalize
Sync Server time with amazon ntp
You have to check by trying to access your bucket and adding file to it after each and every sync with following servers..its like hit and try depending on your region.
ntpdate -u 0.amazon.pool.ntp.org
ntpdate -d 0.amazon.pool.ntp.org
ntpdate -u 1.amazon.pool.ntp.org
ntpdate -d 1.amazon.pool.ntp.org
ntpdate -u 2.amazon.pool.ntp.org
ntpdate -d 2.amazon.pool.ntp.org
ntpdate -u 3.amazon.pool.ntp.org
ntpdate -d 3.amazon.pool.ntp.org
That's it that solved my issue out
