Hello,
I've been tasked with trying to figure out a backup strategy with my web module.. essentially this is what it does...
I have a tag library that works as a resource localization engine. Internally there's a xml parser that reads in xml files from a configured file/URL location on the webserver. Works great for reading. I can access the files no problem.
Now there has been a requirement that is requesting this service maintain backups or "known good" copies of these xml files as they are successfully read.
Basically they're trying to minimize the damage if the service for deploying the xml files should fail and/or the original files become corrupted.
So the process would ideally work like this:
- At time interval attempt to get/parse the actual file(s).
- If succeed, take valid file, and copy it to a "known good" location. (in order for this to work at all, the initial read has to be successful or initial drop of files must be made to both original and backup locations).
- If there's a read failure on an actual file, load from "known good" location.
After realizing that normal file writing IO methods won't work in the container for my purpose I'm wondering what other options I may have for creating the backup files?
I really don't want to go through the process of configuration a datasource or some other heavy handed option... maybe copying the information to a published url? But of course this has its inherited latency problems because of the wire.
I'd be very appreciative of any suggestions of relatively easy approaches for this matter.
(frankly I think this is a matter of over engineering for something that shouldn't happen, and if it does, it should fail and get fixed in real time)
regards,
K2
