The script just sounds like basically what logwatch/fail2ban and maybe a version of DDoS-Deflate.
What I don't understand is what the script is about though. You state you have DDoS Protection capacity, however you also state that it's server-side/host-node side.
The issue here is there is no consistency. The support tech didn't read through the previous tickets to make sure everyone was on the same page. One person is talking about installing a script server-side for mitigation (with the word "beta" involved), another person is talking about their protection capacity switch-side.
What exactly is the uplink/port-speed of your host node? If we talk Swiftway, that's 1 Gbit port speed unless you're paying an additional 400 dollars for 10 Gbit (which I doubt). Similar situation with Psychz (however 10 Gbit isn't as expensive as it is with Swiftway).
If you're talking about additional mitigation strategies after it's already gone through Psychz's/Swiftway's filters, then I think that's very redundant and the entire fact that certain capacity of the malicious traffic making it server-side (since I'm assuming we're talking DDoS/DoS in terms of volume and not something like Layer 7) defeats the entire purpose of server-side protection. By then, it's probably just easier to black hole the IP temporarily.
If
@tr1cky has any monitoring setup and can show us the bandwidth usage ramping up and then cutting out due to "DDoS" (because this topic suddenly went from "Oh they moved people randomly across the pond then back" to "Your VPS was getting DDoSed."), that'd be appreciated. Since the very nature of DDoS is distributed and therefore usually "ramps up" in bandwidth usage rather than hitting all at once.
It wouldn't be the first time someone is accused of getting DDoSed or sending out a DDoS due to misconfiguration in the Provider's monitoring/switch settings. Entire reason why I highly recommend remote server monitoring as well and to compare the data you have with your provider (with the proper time-step of course). Most of my remote monitoring services operates at 30 seconds to 1 minute time-step and has helped me many times when talking with the provider.
I don't know what actually happened. This is all he-said/she-said, however it kinda sounds like snake oil.
Anyone else find their "Beta DDos Protection script" a bit redundant and pointless? I mean if the DDoS attack is more than 5 Gbit (which I mean by itself is a bit hard to get unless you really pissed someone off), then an extra 1 Gbit port saturation is redundant because they'd rather black hole your IP anyways to prevent other clients from being affected.