(Updated)
For the past three four days we’ve been witnessing a great example for the power of community. Zooomr, “The Little Photo Sharing Site That Could… ” has been down for 12 days now. Unfortunately, not for the first time
A few months ago after a failed attempt to upgrade to a major new release, Mark III, Zooomr had to roll back to the previous release, and the whole affair cost two days of outage. In fact the service has never been particularly reliable, temporary outages and slow-downs are quite common. (I know this first-hand, whenever they go down, my blog looks ugly with just placeholders instead of images.)
Then, on March 21st TechCrunch announced the Launch of Zooomr Mark III, with a better look and a host of new features. This post turned out to be premature: the launch failed. Readers who clicked through found a dead site – instead of the Zooomr they could watch Founder Kristopher Tate on Ustream.tv, and some blurb on the Zooomr blog about technical difficulties. (I still can’t decide what’s more exciting: Kristopher sleeping live or Justin.tv doing the same…) On a side-note, this may very well be TechCrunch’s worst blunder ever, and uncharacteristically of Mike, no correction has been posted ever since.
Days later Zooomr was still off-line. Failing twice so spectacularly, along with the smaller problems would normally be enough to bury any startup – except Zooomr, which has a cult-like fan-base undeterred by anything. Is it the product features? Or the fact that Kris was 17 when he started working on it? I can’t figure out the magic, I just sense its presence.
On the eighth day after the TechCrunch announcement, Zooomr Mark III came up live – for a grand total of 10-15 minutes, when their database server crashed. This appeared to be the last straw… Thomas and Kris realized they needed help. The Zooomr site now has a button to make a donation, and Robert Scoble posted a call for help. Zooomr users, at least the more vocal ones still did not lose their faith; they kept on cheering Kristopher, thanking him for his heroic effort… it looks like in the Zooomr world trying hard is good enough… Wednesday morning Zooomr’s PayPal donation account was over $1,500 – some chipped in $5, others $20-30. “Normal” startups would kill to have such dedicated fans/users – I don’t know how Zooomr achieved such status, but any service with such a fanatic fan-base is worth saving… and the saviors arrived soon: first Zoho, then Sun.
Zoho offered its data center and and an additional server to replace the dead one, with identical configuration. Zoho’s Raju Vegesna spent the entire day with the Zooomr team, moving/installing their gear, and configuring the new server. Users and the world could follow the entire process on Ustream.tv. Then word got out that Sun offered one of their “big irons” a Thumper for a 60-day loan. By midnight all the servers were configured, the day ended on a positive note, which is certainly reflected in the flood of enthusiastic thank-you letters Zooomr users sent to Zoho:
Subject :You guys ROCK!
================ Forwarded Mail ===============
You’re willingness to lend a helping hand to Kris and Thomas of Zooomr is simply awesome!
As a member of the Zooomr community, I say thank you and YOU GUYS ROCK!
Subject :Way to go!
================ Forwarded Mail ===============
I wanted to send you this email, saluting your efforts to help a fellow Web 2.0 visionary organization, Zooomr.
I am a Zooomr user, and a proud member of the ZooomrNation. With their recent server problems, I knew that only a select few would even entertain a conversation with the struggling Zooomr crew.
So, thanks. Thanks for your help in our time of need. You are our Samaritan.
[name removed], proud new user of Zoho.com
The last letter shows Zoho may have picked up a few users in the process – well, I’ve said before, you don’t have to be entirely altruistic to do good.
If Day 1 of the crisis (day 10 counting from the failed Launch) brought progress, Day 2 turned out to be a soap opera.
No status update anywhere on Zooomr’ site or blog: the only “information” available is watching Kristopher Tate sleep live on Ustream.tv. OK, he needs to sleep.. but when he wakes up around 11am, he spends another hour in bed chatting on Ustream.tv. Dude, your system is down, but now you have all the hardware you need get out of bed, do something!
By the time he finally arrives to the data center, the Sun team is there with the Thumper. It turns out their machine needs 240V power and the site only has 110V. Phone calls, consultation… Sun finally says they can plug it in anyway, even though it’s not officially supported. Oops… it draws way too much power. More phone calls… finally the Sun team pulls out half the drives and declares the server is safe to use this way. A few minutes after they leave, the server goes down. If these comments are half right, better be careful guys, you may be liable for a crippled Sun!
Never mind the Sun server, it turns out the dead original server may not be so dead after all. Scoble, be proud, Michael Dell is reading your blog! He personally emailed a Dell tech to get over to Zooomr… and voila’ – by late evening the original “dead” Dell server is fixed. In the meantime Kris chats with the Ustream crowd, telling them he intends to bring Zooomr back up on the fixed Dell server.
But by now, two days into the crisis everyone wants to help: a Microsoft Evangelist says “call anytime”. Late evening Kris and Scoble chat on Ustream about how to contact this MS guy. But more advice is the last thing Kristopher needs – all he needs is focus on getting the system back online. Late evening he rattles off a number of hurdles he still needs to overcome, listing the power consumption problem as the main one. This obsession with the Sun machine is becoming a distraction. More about that tomorrow…
Day 3 (day 12 offline) early morning finds Kristopher in the data center – looks like he spent the night there, and there’s a Sun rep with him. Heroic effort, but why? He must feel like a kid in a candy store… wants to have all those goodies around him – understandable, but that should not be the priority right now. He’s had all the hardware he needed since Wednesday night, why isn’t he focusing on getting the service back online?
I’ve never thought I would agree with Shelley Powers one day – she often attacks people and tends to be mean. Her comments on Scoble’s blog were somewhat vicious… but I have to admit she raises valid points. Zooomr is a great service (when it runs) but is far from being a professionally managed company, as recognized CEO Thomas Hawk himself. Forget all previous blunders, let’s just focus on the current crisis:
- No communication to users for days – *CEO* MIA for the second day. “Cool” (childish ?) videos instead of facts.
- No clear action plan to bring to service back.
- Fiddling around with additional equipment that’s a nice-to-have but not needed to bring Zooomr online.
I am not questioning Kristopher’s good intentions. He is thinking of the “bigger picture” – wants all the server capacity he can have, upgrade to the new system with all the bells & whistles, fix some of their scalability issues (MySQL vs. PostgreSQL)… you name it. He wants it all perfect, and perhaps thinks now at day 12 time really does not matter anymore .. just do everything right, however long it takes.
Good intentions aside, what he really needs now is razor-sharp focus on doing whatever it takes to bring Zooomr online now. When you run a Software-as-a-Service business, even if it’s all free, people, in this case 50-100,000 users become dependent on you – that’s a responsibility. Understanding that responsibility is what differentiates business leaders from dreamers – however well-intentioned, talented they are.
Don’t get me wrong, I still like Zooomr, the service, if and when it comes back. I like it mostly for the features (some of which found their way into Flickr) and the attention they paid to bloggers when they gave us all Pro accounts in the early days. Other users will stick around for the community – arguably the biggest value Zooomr has now.
It’s for this community, and the 50-100,000 users (different numbers float around) that I thank Zoho, Sun and the numerous individual users who donated, for stepping up to help save Zooomr. The “saving” part is done. I’ve been updating this article for three days now – I am hitting “post” in the hope there will not be a Day 4 (13), because we’ll see Zooomr live today. It’s up to Kristopher now.
Update (6/2): Day 4 arrived and Zooomr is still down, but there are signs of life coming back. Static photos were back up yesterday (older blog posts no longer show a placeholder for images), and the Zooomr blog is live again. Robert Scoble drove Kristopher back to the Zoho data center for another long night session, and Zoho’s Raju Vegesna also arrived. At about 2am it looked like all the hardware issues were resolved, and Zooomr may be restarted later today – after Kristopher gets some sleep.
Update (6/2, 11:08PM): Zooomr Mark III Launched just this minute. It appears to be working.11:25PM: It’s down. 11:35PM It’s up. Well, I’m not going to declare this a Launch until at least tomorrow…
Update (6/3): Check out the Zoomr and Zoho blogs for updates.
Update (6/3): The Zooomr Saga Continues
Update (7/25): Sleeping on the job.
Tags: zoomr, zooomr, zoomer, photo sharing, online photos, flickr, zoho, sun, sun micro, thumper, dell, web 2.0, community, web 2.0 community, samaritan, goodwill, startups, entrepreneurship, ustream.tv, scobleizer, SaaS, business responsibility
Recent Comments