This is the third installment of a mulitpart part series where we aim to share with you some of the technical aspects of what powers the Managed WordPress Hosting system we developed here at page.ly, how we started, the recent server improvements and a bit on the things to come. [Part 1] [Part 2]
Page.ly finds a new home and begins to scale.
Mid 2010 and Page.ly was really starting to take off. It was time to think about scaling and we also needed a new hosting partner that would help us manage the hardware. We were also very aware of security concerns and sought out a partner who could help.
Firehost.com met all the criteria, I knew the CEO, they were small (at the time) and nimble like us, and security was their core offering. We made the choice and over 1 evening migrated our single server over Firehost.
Firehost has proven to be such an asset to us that we cannot recommend them enough. Apparently others agree as 1 of the first things our early competitors did when setting up was beeline to Firehost as well.
Page.ly also saw a couple design refreshes during this time.
The move to Firehost added multiple levels of security to our system. It was a huge addition to our offering and a value add to our customers as well. We pivoted our marketing a bit to stress this new benefit and have not strayed from it since. We all saw how the big econo-hosts faired in the summer of 2010 as they all got railed by malware. There is no profit-margin on $3.99 hosting to include security I guess.
New Security features of v2 included:
- Managed Redundant Firewall Protection
- Managed Redundant Web Application FireWalls
- Managed Redundant DoS/DDoS Mitigation
- Multi-level Intrusion Prevention and Detection
- Real time virus scanning
- Real time malware scanning
Now here I have to say in hindsight we made a mistake. Rather then push forward to the stack we will describe in the next part of this series, I elected to just scale out with additional Plesk servers. It worked okay, but it also postponed a ‘proper’ solution. The customers were coming in hot, and we were still a company of just 2 founders. This decision to just replicate our single server solution cost us time and money, as it made it harder to migrate later.
So over the course of the next year we went from 1 server to 4 servers. All large monolithic beasts. 8cores/8-16GB of RAM, each one a single point of failure. We had to monkey with DNS, assigning specific clients to specific IP’s/nodes etc. We still used Plesk to manage things and that made load balancing near impossible.
We at least made the wise choice at this time to offload MySQL to a single dedicated machine. Here again though a 4core 18gb ram beast, another single point of failure. #fail
Hindsight is always 20/20.
Under the hood of Page.ly v2
- All nodes HA (High Availability)
- 4 web servers running Plesk & Litespeed, doing DNS as well
- Single Mysql Node
- Used the Plesk SOAP API for vhost configs
- Started experimenting with WordPress caching plugins
- Opcode Caching
So we were moving in the right direction. We had better security, we had enough power to serve thousands of domains and things were moving right along. Speed wise this setup was not bad. Initial hit (not full page load) times were about 200-400ms. If you were a page.ly customer back then you felt some of the pain though of our single server per user setup as we would have to take down an entire node when we needed to run maintenance on it.
In 2010 we also:
- Created our Reseller system that allowed folks to white label page.ly
- Launched our Affiliate system
- Deployed our Vertical Platform that allowed select partners deeper integration with custom reseller packages
- Kicked email off the system and moved it over to an outside provider.
- Had a PowerUp system for a while packaging bundles of themes and plugins from out providers. We moved away from even though it is was fairly successful.
- Flew pass 1000 customers.
In late 2010, we knew we had to scale again.. well scale better. So we embarked on what would be an all consuming project known as page.ly3.
WordPress has a community, and one filled with egos. We ran into a few of these ego’s along the way. Initially a big part of our strategy was engaging the WordPress community. Unfortunately we were soured a bit on the whole thing. For the better we turned inward to improve our product and pursued alternative channels and marketing strategies to recruit new customers which have been very successful.
The market heats up
In 2010 we saw the arrival of a handful of what most would refer to as competitors, it was inevitable. We did the hard part proving the space had legs and there was revenue to be made. And no good idea goes un-copied for long. We were also starting to make a name for ourselves in the WordPress community and that helped quite a bit to get the word out.
A couple interesting points regarding these new arrivals. At SXSW in 2010 I went to the WordPress BBQ at a co-working space in Austin where I handed out a few shirts and were talking to folks about what Page.ly was. One fellow I distinctly remember talking to must have really been listening, a few months later he was a co-founder of a competing company. Around the time this company was launching their other co-founder solicited us about using our technology to power their new offering rather than ‘re-inventing’ the wheel. I was amicable to the idea (fits squarely with our collaboration over competition philosophy) and agreed to a phone conversation which never took place. They decided to roll their own and off they went.
One of the other new players was not happy merely re-factoring our idea, but went so far as to be heavily ‘inspired’ by our marketing to the point I had to have a private conversation with them about the overt similarities between our website copy.
At the end of the day, we did not then and we still do not see any of these chaps as competitors at all. We work from a mindset that with 30+ million WordPress powered sites out there the space is deep enough to support 5-10 page.ly’s and we rather pull clients over from the econo-hosts like bluehost, mt, and godaddy and provide them with a better product.