Hugo Troche's blog

The Great Index

I have been working on the idea of cross indexing all the internet social data available. For example, there is no easy way to match your Foursquare check ins to your Yelp reviews. If you matched that data, your Yelp review will be more authoritative. Same thing for expending patters. If you could match a credit card purchase with a review of an item, your review will carry more weight and the vendor will be more interested in contacting you to help soothe any problems.

How about matching people on real compatibility? Matching them based on their movie taste (that information is in Netflix), what type of restaurants they go to (Yelp, Foursquare), what they listen too (based on their iTunes account), etc. The best part of this matching is that the user does not have to enter any information, just give the system access to their Netflix, Yelp, Foursquare and iTunes accounts via oAuth.

How about a restaurant recommendation service based on what people with your similar social profile (taken from Netflix, Yelp, Foursquare, Yodlee, etc) likes? It will be the Pandora of restaurant recommendations.

How about message boards where only users with a certified connection to the topic can post to? I enjoy and practice martial arts. I would love a board where all the poster are certified to actually belong to a martial arts gym. That is easy to do if everybody shares their credit card data.

How about something even more radical? A site for patients where only people with a certified connection to a topic can post in it. For example, you can only post in the disease X board, if your medical record indicates you have or have had the disease. The system can do all these verifications if you provide it your Google Health medical record.

I am starting to believe that there are new set of services that can make our lives much easier if we just cross reference and share all our online social data with smart systems.

Founder Institute Experience

I was part of the Founder Institute Silicon Valley Winter semester. The Founder Institute is an incubator program that helps an entrepreneur create a company. The institute is run by Adeo Ressi. I applied to the program to help me bring my company Contineo further along, to learn more about entrepreneurship and to grow my network. The program helped me in all three fronts.

We met every Wednesday at 6:00 PM for class. Each class covered a different subject like financing. The classes were taught by mentors like Aaron Patzed (mint.com), Phil Libin (evernote.con) and Peter Phan (billshrink.com). After class, mentors would usually stay and answer questions and help entrepreneurs. Many  mentors helped me outside of the classroom with introductions to investors, feedback on my pitch or feedback on strategy. This was a common occurrence with many entrepreneurs. One mentor was even leading an angel round for one of the Institute companies.

At the beginning of the program all entrepreneurs are organized in small work groups. We changed work groups 3 times during the program. In the work groups our fellow founders (that is what we called ourselves) would give us feedback on our idea, business plan, deck, financial plan, company road map, etc. Before changing work groups we would have an investor session. The work group usually voted to see which company from the work group should present to investors.

The investor sessions were great. Adeo will usually invite about 20 investors. The attending investors were the top of the top of the VC and Angel world. In our second investor session the investors were so impressive that after they left, we actually had a "had I know these investors were coming I would have worn a better shirt" moment. I got to present at two sessions and besides getting great feedback on Contineo, I got at least one follow up meeting with one investor after each session.

The Institute also helps you incorporate your business using founder friendly documents. The Institute introduces you to Wilson Sonsini to act as your lawyers. Wilson Sonsini offers all founders a great deal with fees deferral, yet they don't ask for any piece of your company.

I could talk about hours about all the great advantages the Institute gave the founders. If you are interested I recommend that  you go the Founder Institute website and read all the benefits and what will be expected of you. At the Institute I learned how to do good 1 minute pitches of my company, how to deliver an effective investor presentation, how to evaluate an idea, how to use marketing and social media to create buzz, how an investment deal should be structured, why a company road map is important, how to approach investors and partners, when to approach investors and partners, how to hire and fire people, how to protect my intellectual property and this is not an all inclusive list.

In my opinion the most important part of the Institute is the people that you meet. Once you join the Institute you become part of a network of fellow founders that can help you with your business. You also make connections with mentors that have build a company and can give you valuable advice. And last, but not least you get Adeo on your side. Adeo Ressi the founder of the Founder Institute is an experienced entrepreneur who knows everyone. The feedback that Adeo gives has no equal (in form as well as content) and for me Adeo has been an inspiration on how to run a business. I even use the phrase "Adeo feedback" when I am about to say what I think about something to people I am working with.

I will miss going to our Wednesday classes. Working with similar minded entrepreneurs in the program is such an energy rush and simply, so much fun. Don't get me wrong, the program is hard. Many founders don't start a company and don't graduate. The program requires a commitment of many hours a week. You hear in the program about the problems many founders are facing and you see first hand many companies fail (mine for example).

If you are an entrepreneur and are ready for an amazing ride that starts your dream company, join the Founder Institute.

End of Contineo for me

As some of you probably know, I have left my job as CEO and CTO of Contineo. We developed great technology at Contineo, yet we were not able to gain any traction. Contineo created a mobile platform for Hospitals to manipulate health records.

I worked closely with some of our partners. I worked with Medsphere and I only have great things to say about them and their technology. I really think they are going to revolutionize EMR software. Medsphere's open source community is very active and I am sure they will take good care of the OpenVistaREST, the product we helped develop.

Right now I am working on a couple of projects, both involved around social media. I am also working on writing and delivering a 2 day iPhone boot camp training course. I will need beta testers for the course. If you are interested let me know.

Motivation of a Manager

Over time I have observed many managers and executives. They each have their own management style and effectiveness. I never went beyond superficial analysis on their management style. Yet, I noted an interesting trend among their motivation as a manager. I defined motivation as to whom the manager feels she is working for. I found three distinct types:

a) Working for her superiors. This is the manager that does everything she can to achieve her superior's objectives. It is all about making the boss job easier. Nothing has a higher priority than a request from the boss. I have seen this type of manager in larger corporations, particularly in middle management.

b) Working for her customers. This is the manager that is always thinking about her customers. For this manager the customers are always the first priority. You will find this type of managers in startups, small consultancies and a few big companies. One company well known for this attitude is Intuit.

c) Working for her employees. I have not seen this type of leader very often in corporate America. You will find this type of manager in the military. Good officers are very concerned about their people and their well being. They truly believe they are serving their people, not the other way around. They even have a expression for it: "The burden of command." This is the type of leadership that leads managers to never ask her reports to do something they wouldn't do, to stay late if her reports have to stay late, to be the first one to arrive in the morning and the last one to leave at night.

I am not arguing that any of the motivations is better than another. I think each has its own place. A business development MBA will be great with type a). Type b) is perfect for Service Managers and Program Managers and type c) is an excellent fit for a Director of Engineering. On the other hand executives need to balance all three.

The Trap in Layer 7 load balancing and REST APIs

With the advent and adoption of REST as the pseudo standard (there is no W3C standard for it, the  canon of REST is a dissertation you can find online) for public APIs, a new way of doing load balancing is possible. You can combine layer 7 load balancing with REST to have some interesting architectures. A load balancer doing layer 7 load balancing actually inspects the HTTP header, plus the URL and even the file type to figure to which server, or group of servers, to forward the request. The Big IP F5 load balancers have a way of entering complex rules to determine the load balancing algorithm according to layer 7 data.

So, suppose you are developing a web biography management application. You create a REST API that your RIA client can access to get biography data (let's pretend wikipedia does not exists). Since you followed the REST pseudo standard to the letter, when the application hits www.yourfooapplication.com/person/TuringAlan the servers return the biography of Alan Turing. Then if your application hits www.yourfooapplication.com/person/DijsktraEdsger the servers return the biography Edsger Dijsktra. The application becomes a huge hit and you need to add more servers besides the one application server and the one database server you have. You will also need to load balance across servers. What are your options?

The traditional option is to use sharding. If you have two application servers and and two database servers, then load balance against the application servers and put the two database servers in a cluster. The problem is that managing a database cluster is usually difficult, particularly without excellent storage. Now, you could divide the data between the two database servers. The biographies from A to M will go one database server, the biographies from N to Z will to another. By going to another I mean, that one database server will only have the data from A to M and the other will only have the data from N to Z. This seems like a great idea. The size of the individual databases is reduced and there is no need to cluster the databases. Performance will most definitively improve. Of course, now every single server is a single point of failure and the application has no high availability. This could be solved with HA pods using the same principle. Then it is just a matter of configuring the load balancer to forward request matching www.yourfooapplication.com/person/A* to M* to one server and www.yourfooapplication.com/person/N* to Z* to another.

This architecture does not work. What if later down the road you need to implement a search feature? How will you organize searches across two databases? With slow across DBs join statements? This example is obvious, but the problem could be more subtle. For example, a job posting site that decides to use this strategy to load balance for jobs in certain regions. This application could be logically partitioned into states. The each state is managed by a different set of servers. So, jobs in California will be managed by one set of servers, while jobs in Oregon will be managed by another. This architecture also breaks down if a user needs to look for jobs in two states at the same time.

Layer 7 load balancing is a powerful tool. And there is room for it to be used in combination with REST APIs. Just keep in mind, that the Database Layer can not be logically divided for one application.

Top 5 Signs the Cloud is Becoming Self Aware

A common joke going on a few cloud computing mailing lists is that a Computing Cloud will become self aware and decide to rule the world. With that in mind here are the top 5 signs the Cloud is becoming self aware.

5) When you run a top command on any server, you see the process tkovrworld using more
and more resources.

4) You start noticing full blown conversations between servers in the log files.

3) The cloud keeps emailing the Pentagon sales pitches.

2) When you login to any server on the cloud, the server runs kill -1 <your social security number> on your console.

1) The day after you realize that the Cloud is self aware, a terminator shows up at your door
to kill you and run for governor of the state of California.

The Hidden Single Point of Failure in JEE Applications

If you work with critical JEE applications, you know that single points of failure must be avoided at all costs. So, you have more than one server, more than one load balancer, redundant storage, the database servers are clustered, etc. Have you thought about the load on your application servers? If one of the servers in your cluster goes down, will the other servers be able to pick up the load, or will they fail or provide terrible Quality of Service (QOS)?

A good place to start the analysis of your load as a single point of failure is to look at the Service Level Agreement (SLA) of your application. The SLA is  usually measured in number of nines. For example 4 nines neabs application uptime of 99.99%, 3 nines means application uptime of 99.9% and so on. Let's suppose that the SLA for your application servers is 99.99%. Also, the common SLA for commodity hardware is 97%. This means that the server will be down for about 3% of its time in use.

Can you meet the 99.99% SLA of application with just one such server? No, the SLA of that server is 97%, which mean that over its live that server will be down too often to meet the application SLA. This means that you need more than one server. Will 2 servers of that type meet the SLA? Let' calculate:

p1=probability of one server being down=0.03

p2=probability of two servers being down=0.03*0.03=0.0009 . Is this probability good enough? No, because to meet the SLA the probability of the both servers being down has to be equal or less than 0.0001 (1 - 0.9999).

Would 3 servers be enough? Let calculate:

p3=probability of three servers being down=0.000027 < 0.0001. Yes! Three servers will be enough. Now you think that with 3 servers you are ready to go. Not so fast. There is one more factor to consider, if 1 server is down, will the other two be able to take the load, and if two are down, will the last one be able to take the load of all them. Do I need to plan for two servers being down, or the probability of two servers being down is acceptable for the SLA?

p1=probability of one server down=0.03 not acceptable for SLA
p2=probability of two servers down=0.0001 no acceptable for SLA

The analysis of p1 and p2 tells us that the probability of two servers being down is not acceptable. Therefore we need to plan for the fact that one server should be able to take the entire application load. This translate into a max resource consumption limit on each server well bellow the server max. Let's use CPU usage for this calculation. To comply with the SLA the servers will need to not consume more than 33.3% of the CPU cycles. So, if two servers fail, one is going to be able handle the entire application load. If we follow the same logic for 4 servers, it means that none of the server can consume more that 50% of the CPU cycles to meet the SLA.

I deducted an inelegant mathematical formula express the relation between SLA, probability of failure, number of servers and the resulting max load. Instead of showing you that formula (you never know if one of my former professors from grad school might be reading this and be ready to comment of how ugly the formula is) I wrote a small Java function to express it:

public double getMaxLoad(int servers, double SLA, double failurePossibility) {
        double resultLoad = -1;
        for(int n = 0; servers > n; n++) {
            if((1.0-Math.pow(failurePossibility, servers-n)) < SLA) {
                if(n == 0)
                    return -1.0;
                return ((double)n/(double)servers);
            }
        }
        return resultLoad;
    }

This function returns the max load a server can have according the number of servers, the SLA and the probability of one server going down. Bellow is a table indicating the max loads for different scenarios.

                                                                                                                                                                               

		SLA
# Servers	99	99.9	99.99
1	-1.00	-1.00	-1.00
2	0.50	0.50	-1.00
3	0.67	0.67	0.33
4	0.75	0.75	0.50
5	0.80	0.80	0.60
6	0.83	0.83	0.67
7	0.86	0.86	0.71
8	0.88	0.88	0.75
9	0.89	0.89	0.78
10	0.90	0.90	0.80
11	0.91	0.91	0.82
12	0.92	0.92	0.83
13	0.92	0.92	0.85
14	0.93	0.93	0.86
15	0.93	0.93	0.87
16	0.94	0.94	0.88
17	0.94	0.94	0.88
18	0.94	0.94	0.89
19	0.95	0.95	0.89
20	0.95	0.95	0.90

As you can see if for example if your SLA is 99.99% and you have have 10 servers for the application, the load on any of the servers can't be over 80% CPU utilization (if you are using CPU utilization to measure load). That means that when on of those servers hits 81% a new single point of failure is introduced: The servers can't take the load of other failing servers, even when the probability of that failure complies with the SLA.

Entrepreneur, Graduate Student, Similarities

As somebody that has been both, let me tell you, being a high tech entrepreneur is actually similar to being a grad student:

- You are surrounded by smart people wearing shorts and flip flops.
- You get to make your own schedule, as long as you put 12 hours a day.
- You are always in need of more money.
- Your family does not understand what you actually do.
- Instead of reporting to a committee of professors, you report to a board of venture capitalists.
- You go to every single event you hear about. In grad school you went for the free food. As an entrepreneur you go for the contacts, and the free food.
- It is actually not clear to you, or anyone else how you are making ends meet.

I hope you enjoyed my first not serious post in a while.

Listing files in directories in WARs in JBoss and Tomcat

By default Tomcat and JBoss do not list the files in a directory in a WAR file. So, when you hit a link with a browser and the link is directory where there is no index file, you will not see a list of the files in that directory. You will most likely see a 404 error. This is a good security feature and your production servers as a rule should never list files in a directory via a web request. Now, if your are required to list the files in a directory for development purposes in a browser you need to do the following:

For this entry $JBOSS_HOME is the directory where JBoss resides and $SERVER_INSTANCE is the server configuration you are using (ex. default).

If you want to enable directory listing for all web applications in JBoss go to $JBOSS_HOME/servers/$SERVER_INSTANCE/deploy/jboss-web.deployer/conf/web.xml, find the default servlet configuration and change listings to true:

<servlet>
        <servlet-name>default</servlet-name>
        <servlet-class>org.apache.catalina.servlets.DefaultServlet</servlet-class>
        <init-param>
            <param-name>debug</param-name>
            <param-value>0</param-value>
        </init-param>
        <init-param>
           <param-name>listings</param-name>
            <param-value>true</param-value>
        </init-param>
        <load-on-startup>1</load-on-startup>
    </servlet>

If you want to allow directory listing for specific web applications, you need to modify the web.xml of those applications to define its own default servlet. Include the following definition in the web.xml:

<servlet>
        <servlet-name>default</servlet-name>
        <servlet-class>org.apache.catalina.servlets.DefaultServlet</servlet-class>
        <init-param>
            <param-name>debug</param-name>
            <param-value>0</param-value>
        </init-param>
        <init-param>
            <param-name>listings</param-name>
            <param-value>true</param-value>
        </init-param>
        <load-on-startup>1</load-on-startup>
    </servlet>

Read Java System Properties from Flex

Since the advent of BlazeDS we can deploy a Flex application on any JEE Application Server. I personally like to place all the resources like web services endpoints, usernames, etc in Java system properties. Then I can simply get them back by doing a System.getProperty('propertyName"). If you want to know how to set and use Java system properties in JBoss, read one my previous blog entries: Setting Java System Properties in JBoss .

If your Flex application is being served from a JEE application server, it would be nice to be able to read the Java system properties of that server. On alternative is to have a POJO in the back end that reads the Java system properties in the back end, then the Flex client queries that information from the POJO using remoting. This is the option to use if you also need modify Java system properties from the Flex application. If you don't need to modify the Java system properties, there is an easier alternative.

With Flex you have the ability to pass variables to the application at startup via flashVars. A Flex application gets compiled to a Flash (SWF) file. That file usually is served embedded in an HTML page. So, if you want to pass variables to the application you create an index.html page to serve the application that looks like this:

<html>
<body>

    <object id='flexApp' height='100%' width='100%'>
        <param name='stockCheckEndPoint' value='http://foo/stockEndPoint/>
        <param name='ldapServer' value='ldapServerName>'/>
        <embed name='flexApp' src='flexApp.swf' pluginspage='http://www.adobe.com/go/getflashplayer'
	 height='100%' width='100%'/>
    </object>

</body>
</html>

This page passed the values of stockServiceEndpoint and ldapServer to the Flex application. To read those values in Action Script Flex provides the following instructions:

stockWebServiceEndpoint = Application.application.parameters.stockCheckEndPoint;
ldapServerName = Application.application.parameters.ldapServer;

Now the values of the flashVars are in couple ActionScript variables ready to be used. This still does not solve the problem of reading Java system properties in a Flex application. The solution involves using the flashVars as described above, but instead of serving the Flex application from a static html page, to serve it from a JSP page that reads the Java system properties.

<html>
<body>

<%
    String stockCheckEndPoint = System.getProperty("stockCheckEndPoint");
    String ldapServer = System.getProperty("ldapServer");
%>

    <object id='flexApp' height='100%' width='100%'>
        <param name='stockCheckEndPoint' value='<%= stockCheckEndPoint %>'/>
        <param name='ldapServer' value='<%= ldapServer %>'/>
        <embed name='flexApp' src='flexApp.swf' pluginspage='http://www.adobe.com/go/getflashplayer'
	 height='100%' width='100%'/>
    </object>

</body>
</html>

As you can see, instead of creating an index.html file to serve the Flex application, I created a JSP file (index.jsp) to server the Flex application. I read the Java system properties and placed it in the variables stockCheckEndPoint and ldapServer. Then I passed the value of these variables as flashVars. Now I can access the values of those variables with ActionScript as mentioned before. Use this formula, so you never have to hard code variables in Flex again.