VentureBeat is back

We launched early yesterday morning, but an overwhelming spike in traffic, a subsequent server crash and no fallback combined to shut down the Web site for a whole day.
We’ve learned some lessons.

1) It is hard to launch a start-up. It is surreal to be on this side of the Internet meltdowns, something we covered smugly in our previous role of employed reporter. I recall how we at SiliconBeat wrote about blog search start-up Sphere on the morning of its launch a few months ago, and linked to them — only to discover they’d been stymied by some last-minute bugs. All this traffic hit their site, when they were down, and much of that traffic may not have returned. Tony Conrad, Sphere’s chief exec, at the time, had a very rough morning. Tony, now it’s your turn to chuckle. We feel your pain.

[aditude-amp id="flyingcarpet" targeting='{"env":"staging","page_type":"article","post_id":1766,"post_type":"story","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"business,","session":"D"}']

2) Got to think big. We were paying $30/month for our old SiliconBeat server. So when people suggested we upgrade to a special stand-alone server for VentureBeat, at $100/month, we thought it a prudent move — placing us at the top end of what we thought appropriate for our old traffic at SiliconBeat. But when high-profile bloggers pointed to our site from their blogs (such as Om, Arrington, Primack. Jessica Guynn, among others) yesterday morning, we crashed and never recovered. We didn’t figure out the whole story. But we’ve changed our minds since this morning, and decided to pay $400/month. Be cheap, but not too cheap. And because we’re boot-strapping, this isn’t always an easy call.

3) If something can go wrong, it will. Murphy’s Law. Toni Schneider at WordPress warned us about this a couple of months ago. He even stood by, ready to host us on WordPress servers for free. Still, Murphy’s law kicked in. We had nailed everything else. Thor had crafted the site, I’d proofed it. But we didn’t take the servers seriously enough. We launched, it felt great to be in the air — “until the wing fell off,” as Thor, my developer put it.

4) Have friends. A dead server makes one panic, and we want to thank the folks for their support, in particular Nik Cubrilovic, over at TechCrunch, for lending sound advice. Read this post he shared with us about the overload earlier this year at Techcrunch.

5) The fight goes on. No point whimpering, or pointing fingers about blame. The next day is coming, and you got to try flying again. We remember Tony Conrad’s smile on his face a week after his botched launch, when everything was going dandy for him again.