Only two comments. I don't know why the "without going past" clause is in there, but whatever.
That's been changed. Minutes after the original post.
Secondly, you will have about 20,000 posters posting multiple times trying to post in a matter of minutes. I'm not a techie, but I can imagine it crashing the site.
I doubt it, as the equipment has often faced similar workloads and dealt with them flawlessly. As I mentioned in a post earlier this morning, anytime the ISP's connection would drop for a minute or so (remember when that was a frequent occurrence and lasted even longer?), when the connection would come up, the webserver would get a flood of request that'd been queueing up, and would subsequently flood the db similarly, but never reach 100% utilization dealing with it.
I think I've got the queue set to 1k requests, which is a workload this equipment can power through in a second or two. We get quite a bit of volume during market hours, especially the last half hour, and the queue is always empty. The webserver and db server each have a pair of Xeon's in them, which makes for 4 virtual processors, meaning, in theory, that they can handle 4 requests at exactly the same millisecond (which would be a rare occurrence) and the vast majority of transactions at the db level (at least 95%) take less than 1ms turnaround time.