First and foremost…that guy in the picture looks stressssssed out!
Now you may be wondering why the gentleman in the picture (above) is in such dire straits (and no, I’m not talking about the legendary band from the 80′s). It may be because someone told him to look frustrated in order to take some excellent stock photos…or for the sake of this blog post let’s imagine he’s you (yup, that’s right) right after you find out your site/app (or a section of it) has been down for the past hour and you have no way of telling your customers. No one’s having fun in this scenario
But what could you have done to prevent this?
Well there’s a whole bunch of things…but we’re really only here to talk about status pages, so I’m going to focus on that.
It’s well known that compliance in the cloud is a huge issue. In light of recent events such as news regarding the NSA and the constant battle of online privacy, everyone was looking for a way to know who has access to their cloud accounts. This is why Amazon created CloudTrail – a web service that records AWS API calls for your account and delivers log files directly to you. That’s all fine and dandy, but what do you do with the data AWS is sending you?
My AWS bill is WHAT?! #FiveWordTechHorrors
— Copper.io (@copperio) December 11, 2013
Here in tech we all have our own unique phrase that makes us tick. A short set of words that has an uncanny way of awakening a long awaited fear. The hashtag #FiveWordTechHorrors truly succeeded in capturing this fear in a beautiful (and very hilarious) way.
As the holidays approach, here at Copper.io we understand that monitoring your business is a priority, especially when there are less people in your organization paying attention to it. That’s why we’ll be here for you, monitoring your apps and ready to help in the event something does go wrong.
Summary: We posted about Stackdock on Hacknernews at 10am PT, and got to number 1 for over a day. Resulting in 35k visitors, over 1,500 signups, nearly 650 decks created, 267 instances run, and 200 new paying customers. We launched no where else — and the search engine result is that for ‘docker hosting‘ Stackdock is now number 1 and the blog post is number 4. Overall – wow – we are ecstatic – thanks HN!
Figuring out what’s going to be popular is, at best, an imperfect science. Despite all the preparation and analysis you can put in, there’s no guarantee that something will be popular. Sometimes you can achieve it, but it takes either luck or good planning to get results.
Two weeks ago, we launched a new product called Stackdock, a service which adds extra options to Docker hosting, and wrote about it on the blog. Since we knew Hacker News is interested in both cloud and development, we posted it there as we felt people there would like to know about.
What happened after it surprised us. After only a few hours up there, it was the top story on the site and the post got over 35,000 views in the space of 30 hours. What made this better was that the only promotion done was on Hacker News itself, we didn’t pitch it anywhere else.
While it’s very difficult to make something popular (or go viral if you want to use a different term), there are a few common factors that helped it reach that position. Here’s everything that happened before and after we posted on Hacker News.
The first thing, and is probably the most important factor in this, is that we know the audience that visit Hacker News. The majority of people who visit there are developers, programmers and those interested in the digital space.
Such a product would interest them since Docker is a popular open source application that allows you to create and run any application. StackDock simplifies this process and builds upon it so, naturally, visitors were going to be curious about it.
The second thing was what we posted and when. Since you have limited space to describe what your article or post is about, you need to make sure it’s both interesting and honest. We felt that mentioning what it is and the cost would be enough to showcase the message. We also posted it at 10am Pacific time, meaning that it was early afternoon on the East Coast and early evening in Europe – so most time zones got to see it while it was fresh.
As we mentioned, the post was a hit with the HN community and reached the top spot. Interestingly it also ranked ahead of a story on Twitter’s IPO, showing just how popular it was with readers!!
Not only did the post get a great score of 364 points on Hacker News, the majority of the 126 comments were positive and also discussing hosting strategies as well and discussing Docker in general.
Its popularity meant that traffic spiked on the blog during these two days. Some of the major stats included:
Another benefit from this is that when you search for ‘docker hosting‘, Stackdock now appears first AND the blog post is fourth. None of us would have guessed that the SEO bump from HN would be so great.
Hackernews is great – for this type of product. Time of day probably helped. We didn’t attempt to game the system – we just posted (and since we we’re in Ireland at the time – went for a pint of Guinness after posting, not expecting such a response!
TL;DR: No price reductions today. More regions hinted at but not announced. Introduced CloudTrail – an audit log (awesome – we’ll be integrating this with Twist and Cloudvertical) that will help enterprise adoption and compliance. AWS is now getting into Desktop Virtualization and App Streaming (which I think goes hand in hand with this) in an effort to extract more margin points from Compute and also beat down the last remaining component of Vmware that they don’t already compete with. And IBM tried to do some guerrilla marketing which Andy Jassy (the man who really runs the Internet) turned into a stick to beat them with!
Amazon is Massive
Apparently it’s now got 5 times more infrastructure deployed than the next 14 competitors combined! I’d take that stat with a pinch of salt, but it’s a fair barometer of just how far AWS is pulling ahead.
Private Cloud = [Still] Evil
IBM running buses around Las Vegas – like Marc Benioff used to do with Salesforce ‘protests’ and cabs (check the link for some good ones!) at SAP conferences – saying that they host 30% more of the top web sites than anyone in the world. I’d love to see some backup to that stat… in fairness if it was ‘IBM Power’ it would be easy to believe – between servers and data centres – but the IBM Cloud – even with Softlayer – hard to see it. Andy to his credit pulled the image up on the stage and used it to berate ‘old guard IT companies’ for spreading FUD. Kudos for that quick presentation edit!
Amazon of course are still pushing the ‘Private Cloud isn’t Cloud’ message – but it’s okay to build a hybrid from your existing On-Premise – naturally using Direct Connect, VPC and Storage Gateway. I think the market still has to figure out what’s real in Private/Hybrid with all the vendors pushing their own biases.
CloudTrail launched – compliance and audit logging – this is a big deal!
This is awesome! Especially for the Enterprise. Compliance is a big issue in the Cloud — and not well served. Finding out who is doing what (not just tags… actual activity streams), setting and enforcing policies around what locations, server types and costs can be. At Copper, we already provide a stream of AWS Events from Cost summaries and budget alerts, including a deep Trusted Advisor integration, but this lets us go so much further. Now we can build a ‘Compliance’ tab and show you an audit log of what users have completed which actions – and let you see alerts when unauthorized actions occur – or are attempted (like someone sending data outside the US; or someone shutting down instances that shouldn’t be shut down).
Trusted Advisor has has helped save 140m through 700k recommendations.
Trusted Advisor is interesting – and we’re a fan and partner (the slide above is playing in the AWS TA booth today!). Trusted Advisor in combination with CloudTrail presents a really compelling activity stream. We’ve already integrated Trusted Advisor events (and we summarise key changes so you don’t have to look through all recommendations every time you look at it) and we’ll be adding CloudTrail events too.
AWS get into Desktop Virtualization
AWS is now getting into Desktop Virtualization and App Streaming. This has long been a stalwart revenue generator for Vmware, Citrix and the System Integrators and Services Providers that deploy and maintain installations. It could be really disruptive. My first reaction was that this is just another way to sell more compute and extract some more margin from it (like RDS and Transcoder – compute with ‘extras’) – but when App Streaming was announced a clearer picture of the plan emerged –> AWS now want to host USERS apps, data, and desktops – not just power web apps. If it takes off this could be a huge revenue driver from AWS.
Finally – the 6 stages and drivers of Cloud Adoption
These are relatively obvious and we see them everyday with our customers – but I thought the 6 stages/use cases were were summarised so I’d repost here:
The AWS blog has tons more detailed info on all these announcements too.
Hope that’s useful
Time flies – especially when something good is happening! Day 2 at RubyConf is history now – but what a great day!
Yesterday the opening keynote “My KidsRuby Journey” by PJ and his daughter Katie – and it blew everyones mind – she nailed it – Matz said that we don’t know our future and it’s hard to disagree but we could see a little bit yesterday how the future might look like thanks to Katie! Oh and one more thing – if you were wondering if kids can code if you were on yesterdays keynote you already know the answer…
Love this photo by PJ
I’ve always had a weird interest in compilers … Tom did great job explaining how all that works and that …
This quote from Compilers For Free says it all:
This math-free talk uses Ruby to explain how partial evaluation works, how it can be used to make programs go faster, and how it compares to ideas like currying and partial application from the world of functional programming. It then investigates what happens when you run a partial evaluator on itself, and reveals some surprising results about how these techniques can be used to automatically generate compilers instead of writing them from scratch.
Next on our list was talk by Stephen Henrie “A Lightweight SOA Framework using Ruby, Apache Thrift and AMQP” and it was good to see Thrift in action but I think the decisions we’ve made at Copper to stick with simple JSON are good ones.
when was saying he is going to RubyConf to give a talk about UML! It was great talk – for all of you missed it make sure you’d watch it once the talk gets online. It’s a must see. (We’ll update this blog with links to Slideshare and video streams as they go up…)
We liked this one a lot…
“If you make your diagrams too pretty you might be inclined to keep them too long”
We’ve also went to “Preferring Object-Orientation to Metaprogramming” by Steven Harms. Steven shared with us how he changed his mind about meta-programming since his last talk on RubyConf 2011 pointing out when to use meta-programming (almost never!) and that you would be better off using OOP best practices in Ruby where Sandi Metz book should be your bible.
My favourite qoute:
“Metaprogramming is therefore… Writing bugs that have adventitious, beneficial side-effects”
Looking forward to another great day today!
Dominik & Kamil, Team Copper.
Yesterday was Day 1 at RubyConf in Miami and it was a fantastic day!
It started with Matz keynote
Matz presentation focused on 2 things – one was the “fantasyland” we were/are living in – that we pretty often don’t know what we are building (but still pretty often we ignore our ignorance) as it’s not physical like buildings, and at the same time we don’t really know what we want and we don’t know the future.
Secondly Matz pointed out a pretty important thing:
“What we need is Garbage Collectors not memory management, but waste. We rely on them to get the job done.”
“A programmer is a creator. Come join open source software as a contributor. Make the world better. Be a Garbage Collector take a part in keeping up the fantasy land. The fantasy land facade is a direct result of the efforts of Garbage Collectors.”
Since at copper.io we have loads of Gems(here and here) for different products we went to two great talks (our favourite statement was that designing a ruby gem is like UX design –> we have to care about users) ‘Extending Gems – Patterns and Anti-Patterns of Making Your Gem Pluggable’ by Jason R Clark and ‘API design for gem authors (and users)’ by Emily Stolfo
We’ve also enjoyed the Zombies talk (‘Fault Tolerant Data: Surviving the Zombie Apocalypse’ by Casey Rosenthal).
The last talk of the first day at rubyconf was ‘Beneath The Surface: Harnessing the True Power of Regular Expressions‘ in Ruby by Nell Shamrell. She not only showed how to write regular expressions, but explained how they are processed and how they work under the hood (Quantifiers and Finite State Machines). Insightful, inspirational, great talk. Thanks Nell!!
And then we capped it off – many thanks to Kickstarter for hosting a meetup and drinks at the Segafredo l’Originale bar!
P.S. For all of you asking about our Copper T-Shirts
– get in touch, with you postal address, and we will get them shipped to you after the conference.
Looking forward to another great day today!
Dominik & Kamil, Team Copper.
We’d a lot of activity yesterday! We launched Stackdock and it was the number 1 post on Hackernews for most of the day – ahead of Twitter’s IPO? Yowza I didn’t see that coming! I guess doing stuff – with Docker – is more interesting to developers than talking about doing stuff!
In the mean time – no doubt some of you are at RubyConf in Miami for the next 3 days. We’re there too – and we’d love to chat – tweet or mail us. No agenda – just want to have as many great conversations as possible.
Anyone going to AWS ReInvent next week – we’re throwing a little party with some of our friends – register for our free Exosphere event in the Palazzo on Wednesday evening. We’re keeping the details a secret for now – but it’ll be fun – and we’ll SMS you on the day with final plans!
Thanks again for all the feedback on Stackdock over the last day. And for my part – I’m really proud of what our team is doing in Copper and our set of tools. So – thanks Coppernauts (yeah, need a better name)!