brendan's questions

I am taking periodic snapshots of a 1TB EBS (Amazon Web Services Elastic Block Store) volume as backup. In the case of the whole AZ (Availability Zone) becoming unavailable, my Disaster Recovery plan is to create a new EBS volume from the latest snapshot in another AZ in the same region.

How can I figure out how long it will take to create the new EBS volume? I have an RTO (Recovery Time Objective) of 6 hours. Can I meet it with this approach?

It probably shouldn't/doesn't make any difference, but I am in the ap-southeast-2 region (i.e. Sydney).

I noticed that in my Apache log (actually Oracle's branded HTTP Server, which is really Apache) the timestamp of the lines sometimes goes down, instead of up. Here is an example:

1.2.3.4 - - [16/Dec/2014:**13:22:46** +1100] "POST /cas/v1/tickets/TGT-4830717-K5FWqTeJfIajdbae0Zbc1hqSOREWG92fybtLKYtdufKve16Uou-cas-1.2.3.4-2 HTTP/1.1" 200 60 
1.2.3.4 - - [16/Dec/2014:**13:12:42** +1100] "POST /self-service/api/account/verification/residential HTTP/1.1" 400 0 
1.2.3.4 - - [16/Dec/2014:13:12:15 +1100] "GET /self-service/api/accounts/38909 HTTP/1.1" 404 0 
1.2.3.4 - - [16/Dec/2014:13:20:46 +1100] "GET /self-service/oamCallback?authn_try_count=1&contextType=external&username=string&contextValue=%2Foam&password=sercure_string&challenge_url=https%3A%2F%2Fwww.xxx.com.au%2Fmyaccount%2FoamCallback&request_id=-3108004070152329208&p_error_code=OAM-2&p_error_codes_list=OAM-2&locale=en_US&resource_url=https%253A%252F%252Fwww.xxx.com.au%252Fmyaccount%252Fapi%252Fprofile HTTP/1.1" 401 70 
1.2.3.4 - - [16/Dec/2014:13:22:46 +1100] "GET /cas/proxyValidate?&ticket=ST-4830713-csoZibAECToKVXXYTIVS-cas-1.2.3.4-2&service=http%3A%2F%2F1.2.3.4%3A7002%2Fservlet%2FREST%2Fsites%2FxxxQuotes%2Ftypes%2FPage%2Fassets%2F1343354161974 HTTP/1.1" 200 704 

See how line 2 has a time of nearly 10 minutes earlier than line 1?

Is that normal?

Or if not, any idea what causes it?