|
|
|
|
|
|
|
|
|
|
 |
|
|
|
Amy Senior Test Engineer |
|
Craig Quality Assurance Analyst |
|
Matt Team Lead |
|
Chris Team Lead |
|
Ray Senior Test Engineer |
|
hello moto :-D
Monday 6:30 a.m. (Home) - Long Beach, NY (Long Island)
Awakening for another busy week, I shower and prepare for my commute into Manhattan. Prior to catching a train (LIRR) into New York City, I take a few moments to enjoy the view of the ocean from my terrace and gulp down a cappuccino. It's the middle of winter, but it's still humbling to gaze at the sea and relish in its peacefulness. The commute lasts 1 ½ hours, door-to-door, as I am whisked away from my beachside community to the hustle and bustle of the Big Apple.
9:00 a.m. (Client Site) - Manhattan's Financial District
The project I am involved with for the next three days encompasses a major pharmaceutical corporation's client/server application used for the data mining of case studies involving drug interactions. Our goal is to determine if the application has the capacity to support its end-users after its database is merged with a larger data source containing additional case histories. In order to accomplish this goal, we are using Mercury's LoadRunner performance testing tool to automate the application's business processes and simulate the application's end-users running reports.
10:00 a.m. I meet with the business analysts and subject matter experts in order to confirm the necessary business processes that should be tested, as well as formulate a testing strategy. There are eight (8) reports that need to be transformed into automated test scripts within LoadRunner.
12:00 p.m. It's lunchtime and an interesting time to walk around Manhattan. There are throngs of people to look at and avoid as you share the crowded sidewalks. There are scores of places to eat lunch at and this usually invokes much indecision amongst co-workers, since there are too many choices. Today the choice is Chinese cuisine.
5:30 p.m. After lunch I forged ahead in creating the automated test scripts. I am about 10% complete with this first phase of testing. The commute home is always interesting. As the train crosses over the water to Long Beach, it always feels that I've entered a different world or that I am back on a family vacation.
9:00 a.m. (Client Site) - Manhattan's Financial District
As I sip on my tall latte from Starbuck's, whole milk and real sugar (no soy, no syrupy flavors), I continue on my automation effort. The automation tools for performance testing are quite interesting, since they create automated test scripts that implement and emulate the network-based communication conversations between two or more computers. The technology exposes the inner workings of distributed computer environments and allows us to emulate many computers accessing a remote application, while collecting metrics that allow us gauge the application's capacity and end-user satisfaction.
12:00 p.m. Today's lunch choice is Sophie's, a Cuban restaurant that is a staple amongst RTTS employees. I have a Cuban sandwich topped with Sophie's fabulous green sauce.
2:00 p.m. Since I am traveling to Toronto on Thursday, I make the necessary travel arrangements during a break.
5:30 p.m. I travel home having completed 50% of the first phase of testing.
Wednesday 9:00 a.m. (Client Site) - Manhattan's Financial District
After I get my espresso-based fixation out the way, I continue to create the automated test scripts for the client's application. This process typically requires on average ½ of a business day to automate a business transaction. However, I anticipate on completing the majority of this first phase of testing by the close of the business day due to our modular approach to business process automation.
12:00 p.m. The cuisine of the day is Thai. A co-worker and I head over to Lemongrass Grill for some Thai Iced Tea, Duck Choo Chee, and Chicken Massaman.
5:30 p.m. I have completed 88% of the first phase of testing. Seven of the eight business transactions have been transformed into automated test scripts. However, the client is not ready to start testing their application until next week, so I have plans to fly to Toronto, Canada tomorrow morning to focus on another project.
Thursday 8:00 a.m. (New York's LaGuardia Airport) - Flight to Toronto, Canada
I hop on a plane to Toronto, Canada, in order to address another client's need. The customer is re-hosting their application within a different data center, as well as updating its application's infrastructure and needs to determine whether the new environment will support its customer base. The application provides frequent flyer benefits to its customers, such as flight booking and frequent flyer mileage redemption.
12:00 p.m. (Client Site) - Downtown Toronto
After picking up the rental car, I drive to the city of Toronto in order to meet with the developers of the application. The objective of the meeting is to finalize a new testing strategy that will flush out the capacity issues we had uncovered a month earlier.
3:00 p.m. (Client Site) - Toronto Suburbs
After the noontime meeting I drive to the data center where the application is being hosted. There is a meeting with the project's sponsor and team leads in order to get everyone in sync regarding the newly adopted test strategy. After the meeting I set up the testing hardware and software within the data center by plugging the testing devices directly into the network backbone of the data center. This data center is rather large and is a central point within Canada's Internet pipeline. So, it's pretty fascinating to hear the buzz of all the machinery and knowing that these are dollars and cents being processed and corporations being managed.
8:30 (Hotel) - Toronto Suburbs
I check into the hotel, which is rather quaint, since it has a warm fireplace to sit by within the lobby. I have dinner at a local restaurant, Brix Napa Valley Grille, where I have a nice glass of wine and a Californian-inspired meal.
Friday 9:30 a.m. (Client Site) - Toronto Suburbs (Data Center)
I arrive at the data center anticipating a long day, since there are many tests that need to occur and business decisions that need to be made by the end of the day. Although the day was long, there were large breakthroughs that were achieved in regards to the capacity issues we had uncovered in the prior month. Our team ran many incremental tests that were orchestrated to isolate individual components within the application's infrastructure. We validated that the web server farm could handle the anticipated load by subjecting the cluster of three (3) web servers to 70 web page views per second across 1,600 concurrent users. We also determine that each application server could support 400 to 600 concurrent users prior to the need to bring additional application severs on-line. When we attempted to test 1,200 concurrent users in a full end-to-end test (400 users across 3 application servers), we uncovered a failure within a critical component, an XML Gateway, which could not handle the rate of incoming requests for services. Therefore, extensive queuing was occurring that was causing performance degradation for all of its dependent components downstream. 6:00 p.m. There was a status meeting with the project stakeholders that highlighted the findings. Additional tests were devised and system specialists were brought into the project in order to troubleshoot the problematic component.
1:30 a.m. (Hotel) - Toronto Suburbs I arrive back at the hotel after a long but successful day. Significant value was added that enabled our client to address a critical component within their application that, if not addressed, would affect the ability of the application to scale as a function of its concurrent user base, as well as affected the generated revenue and customer satisfaction. After checking-out, I head back to the airport and fly home. The weekend is supposed to be nice. My plans are simple. Sit on my porch, and have a well-deserved rest.
Monday 6:45 a.m. - My house, Queens, New York Rise and shine. It's the
start of another week. I have my usual morning cup of coffee with hazelnut
creamer, get dressed and head over to the subway station. It takes me about 45
minutes each way in rush hour, but at least I am not doing the driving.
8:35 a.m. Client Site, Financial District, Manhattan On my current
assignment, my client is one of the world's leading financial management and
advisory companies. I have been on this project for almost a year now. My team
members from RTTS and I (6 of us in total) have automated the testing of their
trading systems. Because of our success here in New York, my project lead and I
have now moved on to test our clients' systems in London and Tokyo. We spent a
week in London recently kicking-off the engagement and will eventually be
working with a team in Hong Kong, as well. I skim through my Task List to get an
idea of what tasks the day holds for me. Afterwards, I go upstairs to the
cafeteria with some of my team members to grab some breakfast. (This type of
work requires a lot of brain power and I need my energy.)
10:00 a.m. An important part of our automated system is a database which
serves as both our data repository and a front end for users. The database for
Tokyo is still in its development stage, so I spend most of my time designing
forms and making sure whatever backend code that is needed is in place.
1:00 p.m. I head out to pick up some lunch with my project lead, and some
other members of my team. We talk about what we did over the weekend and joke
around until our break is over. This is a reprieve that I can count on everyday,
which is nice. We almost always leave the building to get lunch and some fresh
air.
2:00 p.m. In addition to working on the database, today I am running some
automated test scripts and fixing whatever data/code issues arise in the test
scripts for the remainder of the day.
Tuesday 10:00 a.m. - Client Site It's time for the first of the bi-weekly
conference calls my Project Lead and I have with our clients in the London
office. Since these clients are also playing an integral role in the project in
Tokyo, we go over the current issues concerning the system in Japan - mainly
whatever defects have been entered in the defect tracking system and script
repository. (On this site, we are using Mercury TestDirector.) The defects may
be part of the database, application code, or erroneous data that has been
entered by the user, etc. We ask questions to make sure we have a concrete
understanding of what the problem is.
11:00 a.m. I investigate the defects and implement whatever solutions are
necessary for the rest of the day. A lot of this involves performing the
particular transaction manually, looking into the data logs and comparing the
data values that are being sent by the system under test to those that we have
in the database.
5:30 p.m. I head out to my weekly swimming lesson at a YMCA in the city.
After years of being absolutely petrified of the water, I'm finally putting my
fears behind me. My incentive? An upcoming vacation in Honduras with my
boyfriend.
Wednesday 9:00 a.m. - Client Site I go over some transaction flows with
my project lead. The Tokyo systems have different types of orders that can be
entered and I want to make sure that I have a thorough grasp of the business
rules.
10:00 a.m. Since the systems are down in Tokyo for a few hours I
concentrate on working on the database. I create three new forms and test them
out by entering new records, ensuring that all the necessary fields are
available and checking to see that the associated table gets populated
correctly.
2:00 p.m. The Tokyo systems are now back up so I take the new records that I
have created in the database and incorporate them into various test scripts to
make sure that everything runs properly. This is exacting work that requires
great attention to detail. It takes me the rest of the day to make significant
headway.
Thursday 7:00 a.m. - Client Site It's time for our weekly conference call
with Tokyo. With the time difference between New York and Japan, we have to
arrive at work early to interact with the development team there. We go over the
raised defects and make sure that everyone involved has a clear understanding of
the process flow surrounding the transactions in question. It becomes evident
after a while that our clients don't have a clear grasp of some of the
functionality behind our automated system - mainly how we're going about
checking that our transactions have gone through properly - so we spend some
time explaining how things work in a step-by-step manner.
10:00 a.m. It's the second of the bi-weekly conference calls with London.
(At this point I have spanned the world in time zones.) We are looking to
complete this "shakedown" period of the Japan automation system by the end of
next week and right now this goal seems within reach. After it is rolled out to
offshore system testers, we will be able to turn some of our attention back to
the London systems and making it more robust.
2:30 p.m. I test out some more forms that I have created and write test
scripts that cover various scenarios. I spend the rest of the day writing
backend code for the database and running scripts.
Friday 8:35 a.m. - Client Site I read my emails and look in TestDirector
for clarification on any defects that have been entered. There are no conference
calls scheduled for today so I'll be able to fully concentrate on the tasks at
hand without interruption.
5:30 p.m. We finish testing all of the fixes we have made and drop a new
copy of the databases and code into the Tokyo shared drive. We detail whatever
changes were made in an e-mail and send it out to all necessary parties. This
should take care of all of the defects that were entered and is a good end to
what proved to be a productive week. This weekend I plan to relax at home and
see a movie with friends…and brush up on my Japanese in preparation for that
engagement.
6:53 a.m. - Home, Queens, New York
My commute to this client is relatively easy so I don't have to get up too early to get to this assignment. I live in one of the boroughs of New York City and the client site is in lower Manhattan. I only have to take one subway and walk a block to their building.
8:00 a.m. - Client Site, Financial District, Manhattan
I arrive at the client site fresh from the weekend break. I am at the client site early every day to put some extra time in to the increasing amount of automated tests that are currently being used. Even though my position is primarily manual testing, we are introducing automated testing to the client as time permits.
9:00 a.m. I sort out the test cases I am going to execute for the day and spend almost a full day executing them. As a Quality Assurance Analyst, I am conducting manual tests to test all the functional aspects of the client's application. This application is used for business transactions for a trading application. Due to the application's complexity, issues often arise during the build deployment that are resolved between brief meetings with our team lead and the Project Manager.
4:00 p.m. I attend a daily defect meeting. The entire development and testing team is involved from both RTTS and the client-side. We discuss the defects that we are tracking and formulate a plan of action to take to resolve each one. The progress of the regression testing cycle is also discussed.
Tuesday 9:30 a.m. - Client Site
Team Update Meeting - Tuesday's schedule is different from the rest of the week because I will participate in two more meetings than usual. The first is an update meeting with my team lead and the rest of the manual testing team from RTTS. We discuss the progress made and any issues that were encountered for the week.
2:00 p.m. I also attend a Business Analyst meeting which is attended by the both my team and the client team that works on this application. We are informed about what changes are going to take place with the application and also if any major issues have been encountered.
6:00 p.m. I am going for a long run after work. I ran on the track team in college and don't want to lose my edge. I hope to run in a half-marathon this fall.
Wednesday 8:30 a.m. - Client Site
I continue to work on regression testing. I sometimes work with the developers when they are fixing a defect that I have entered. I will retest the function of the application that was associated with the defect after the correction was made. This is so the developers can see the scenario from the user point of view.
4:00 p.m. - Defect Meeting I attend a daily defect meeting. Wednesday's defect meetings usually have a little different feel to it because Thursday is usually the day that a new version of the application is released, so the client is forming the final opinions of this version. The amount and severity of the defects that are being raised at this point are taken into careful consideration. Also, because the release date is the next day we usually stay a little later to make sure that we execute the entire test in the regression cycle. During the regression test, we selectively re-test the system that has already been modified to ensure that any bugs have been fixed and that no other previously working functions have failed as a result of the reparations and that newly added features have not created problems with previous versions of the software.
Thursday 7:00 a.m. - Client Site
Today is the day of the new release and the version we have been testing goes into production at the end of the day. We have to have all the scripts run by the defect meeting today. I also go over the defects that I have notes and re-test them. Following the re-test, I give each defect a final resolution in the tracking system. Some defects will be closed (meaning they are fixed) and others will be marked to be reopened in the next build.
Friday 8:35 a.m. - Client Site
Friday's schedule depends on whether or not we are starting another regression cycle. If no changes are ready to be put into the QA environment, then we have a time to focus on move from manual testing into automated testing. This is important to us as manual testers because the more test cases we have automated the faster we can progress through the regression cycle. If we are going into another cycle immediately the team leader will reassign the tests and e-mail us our assignments, and then were in the cycle again.
4:00 p.m. - Defect Meeting The defect meeting the day after the release usually includes a report from the client's Project Lead describing how the new version has been received by the users. Ultimately, our goal is to provide a new version that functions the way the end users expect. So far all the releases that I have seen have been successful and I hope that they all will be.
5:30 p.m. Well, the end of another week has come. After work today, some of my RTTS co-workers are going out with a few of the developers. Over the weekend I am going to meet some friends that I know from high school at a club in Gramercy. That should be fun.
Monday 6:30 a.m. (Home) - Suburbs of New Jersey
My commute this week has been slightly disrupted by a transit strike in New York City. My trains from the New Jersey suburbs are more crowded than usual, but otherwise I can still manage to get in to the office. I take an express train and change to the PATH train to traverse under the Hudson River into Manhattan. Later this week I have to fly to the client site in Atlanta so this won't be a long-term problem.
8:50 a.m. (RTTS' Corporate Office and Test Lab)
The current project I am on is a unique assignment. As the Team Lead on this project, I am only responsible for one test engineer. He works from RTTS’ test lab in New York City, while I travel between Atlanta and the test lab. When I arrive, he is already writing and running test scripts.
9:15 a.m.
I join my Division Manager in a meeting with the Director of Professional Services so that we can review the status of the project. The client is very happy with the progress we are making and has asked us to continue testing additional builds of the application. The application is a retail web site hosting more than 3 million listings of products. The primary focus of our testing project is to create an automation suite for the main consumer site application using an industry-leading functional testing product. Eventually, we will transition the automated testing to the client’s internal testing team. I provide statistical data during the meeting representing the number of test cases we have completed and defects that have been reported during the last week. We also discuss my next visit to the client site this week.
11:30 a.m.
I continue to edit test scripts for the rest of the day and prepare the final reports for my client meeting tomorrow. We have VPN access to the application’s most recent builds on the test bed. This allows us to work remotely while maintaining a real-time presence.
Tuesday 5:00 a.m. (Home)
Today I am flying down to work with the client in Atlanta. I get dressed and head to the airport to catch a shuttle flight.
6:45 a.m. (Newark International Airport, NJ)
I am checking in to my flight to Atlanta. Everything is on-time and I should be at the client site later this morning. I am scheduled to brief some of the key players this afternoon.
11:00 a.m. (Client site, Atlanta, GA)
All went well with the flights today. I am in my workspace and have checked-in with the QA Manager here to make sure everything is ready for the meeting. I will review my reports while I grab something to eat in the company cafeteria.
1:00 p.m.
My briefing starts on-time. The Project Manager, Automation Lead, Development Manager and QA Director are all in this meeting. We start by reviewing the metrics of this project. Since the client has asked us to expand the scope of the project, I present the updated timelines. Everyone is amenable. We also review the defect reports.
5:30 p.m. (Buckhead Area, Atlanta, GA)
After checking into my hotel, I meet with the Automation Lead for dinner. He wants to discuss the eventual transition of automation to a team of their own employees. RTTS will provide mentoring and formal training to support this endeavor.
Wednesday (Client site, Atlanta, GA)
I will run scripts from this location today. I use instant messaging to communicate with the Test Engineer testing remotely at RTTS. He has made significant progress writing the scripts. I will be able to report that we are ahead of the estimates in my weekly report on Friday.
12:00 p.m.
It's lunchtime and an interesting time to walk around downtown Atlanta. I grab a sandwich and walk over to Centennial Olympic Park. This is the site of the 1996 Summer Olympic Games. It is amazing to see how they are using these fields and structures today. The hotel I am staying at is on the other side and I have a nice view of the parks and ball fields.
5:30 p.m. After lunch I forged ahead in creating the automated test scripts. I do a final check of email before wrapping up for the day. I have to check both RTTS and client mail throughout the day. I am flying home tonight.
10:45 p.m.
Just landed at Newark. It has been a long day, but I am glad to be almost home.
Thursday 9:00 a.m. (RTTS) - Manhattan's Financial District
The transit strike is over and my commute it back to normal. I stop at Starbuck's for a double espresso. I need the extra dose of caffeine today. While my counterpart works on additional scripting, I am writing documentation for the next 2 days. This document is an important reference document the process we used in order to develop the automation suite, build reports that outline the progress of the testing initiative, along with weekly reports that provide a current status of the project and assets. It is very detailed and will be used by the project management and development teams.
12:30p.m.
Today I am having lunch with some of the Test Engineers from my past assignment. They are still working on that pharmaceutical giant’s site which is close to our offices. It is rewarding to see they are doing well. I like to think that my leadership and experience from my time as their Team Lead contributed to their success and our company’s. There was a lot of camaraderie with this group, especially since we participated in a soccer league after work. We competed with other teams, most of them a lot better than we were, from other companies. We practiced once a week and played a game every Friday night during the season. The games are usually late at night, but we had a blast. Last season, we had official team shirts sponsored by RTTS. Our season record was an admirable 4-6. I hope to play on the team next season, although most of us will be on different client sites by then.
5:30 p.m.
Having made significant progress today with the documentation, I head home to relax and do laundry.
Friday 8:30 a.m. (RTTS' Corporate Office and Test Lab)
It has been a long week but we have accomplished quite a bit. Today I complete the documentation and send it to my Division Manager to review.
12:30 p.m.
Lunch today consists of sashimi. A group of newly hired employees are talking about their projects. I am able to lend one of them a hand with a question he has with one of the automation tools. Everyone here helps each other. Most of the senior folks have been in a variety of situations and this often comes in handy to work through a problem.
5:15 p.m.
The weekend has arrived. A group from the office head down to one of the bars nearby to unwind. We try to visit different places every month or so. Today, a good Irish pub is where we’ll be.
Monday – 6:30 a.m. (Home) – New Rochelle, NY
Monday’s are usually thought of as the least favorite day of the week, but this is no ordinary Monday. Today, I am starting a new engagement at one of RTTS’ most well-known and prestigious clients. This site is the headquarters for one of the world's largest manufacturer, seller and distributor of some of the world's most recognized consumer soft drink brands. The excitement brings forth a slight sense of nervousness, and there is nothing better to settle some butterflies than a slightly frozen, blueberry yogurt protein shake. Trust me, I am not overly health conscious, but I have always been physically active in school and do my best to maintain that level of fitness now that I am working full-time.
8:30 a.m. (Client Site) –New York Suburbs
I arrive at the client site about 30 minutes early. I was anticipating highway traffic, but to my surprise the 35 mile drive was smooth and even somewhat relaxing. I decided to take a parkway because it is a notably scenic route this time of year.
9:30 a.m.
I meet with the Director of Quality Assurance to discuss the current project that has been assigned to the QA team for performance/scalability testing. The application that needs to be tested is Web-based, built upon Java architecture. It will serve as a link between two separate human resource systems and will be used by payroll administrators. In the past, payroll processing time was lengthy because the administrators were required to manually compute various calculations. This new tool is intended to reduce the processing time and allow for them to meet deadlines. During my meeting with the Director of QA, we outline her expectations for testing and the timeline. I will report to her throughout the project and keep her updated on the metrics of the project.
12:00 p.m.
The Director of IT Applications comes by around lunch time to take me on a quick tour of the building and welcome me to the project team. The building’s architecture is exquisite, yet somewhat confusing, especially because the main entrance is on the 6th floor. The testing and development laboratory that I am working in is located on the 4th floor. Taking an elevator down to the 4th floor upon arrival will definitely take some getting used to. We also discuss the project and some of the other applications currently in development. He feels that we can help him again in the future for these short-term performance tests – both to augment his existing team or to run all of the tests on our own.
3:00 p.m.
I meet with the application’s developers and subject matter experts to identify the business transactions that will require automation and clarify the user load expectations the system is required to handle. The primary objective of my testing is to determine if their expectations regarding application performance are achievable when it is tasked with upwards of 100 concurrent users. Our immediate goal will be to demonstrate that the environment will hold up in-production for a population of only 100 users on the system used in Canada. Testing higher scalability for the U.S. release is desired, but isn’t as important as the Canada initiative. Considering this information, I am able to formulate a tentative testing strategy. Time willing, I will execute further testing that implements a user load greater than 100 concurrent users simply for the purpose of generating response times and system capability expectations. This data might be helpful when we move forward with the future release. A secondary objective is to uncover a breaking point of the application – how many users and transactions can it handle before the application fails. I am using Mercury LoadRunner to automate each of the identified business threads.
5:30 p.m.
It is time to head home. By the end of the busy day, I have successfully created sample automated test scripts to test the compatibility of LoadRunner and the application. The tool was able to effectively make the appropriate HTTP requests and correlate required server generated data. The actual number of automated test scripts will be lower than the number of test cases, however. Logic will be implemented throughout the scripts to give flexibility to each virtual user to navigate through each of the desired paths of the application.
Tuesday 9:00 a.m. (Client Site)
During the building tour yesterday, I saw the 5-star cafeteria they have on-site. Because of the rural location of this massive facility, there aren’t too many places to get breakfast or lunch. The cafeteria is incredible, so I decide to grab breakfast there today. After yesterday’s long day, I decide to have a big meal - a bacon, egg and cheese sandwich on a sesame bagel should hold me through the morning.
12:00 p.m.
The statement of work from RTTS’ business development team is a key document that contains a description of the load testing process, allotted time estimations, and final deliverables. I make sure that everyone I will be working with at the client site has this document, so I pass it along. Then I get started automating the required business transactions.
6:00 p.m.
57% of the first phase of testing is complete. By end of day, 4 out of 7 business transactions were successfully automated. All appropriate information had been correlated and data-pooled. Random functionality to select between available payroll accounts, as discussed with the developers, has been successfully implemented throughout my automated test scripts. Functions have been created to successfully parse the server generated session ID from the HTTP Response header, a critical piece of data in properly automating the specified business threads and implementing a concurrent virtual user testing strategy. I am currently working in and creating scripts against the development environment, which points to the production database. This is the only environment that contains the necessary data to complete testing. Testing is scheduled to begin tomorrow after 5:00p.m, assuming that I do not run into any obstacles automating the remaining test cases.
Wednesday 9:00 a.m. (Client Site)
I added some variety to my breakfast this morning and substituted thinly-sliced ham for bacon and two cold Starbucks Vanilla Frappucino coffee drinks. I am able to bring my breakfast to my workspace. This allows me to maximize my time as I continue to work on the test automation effort.
6:00 p.m.
I have completed 100% of the first phase of testing. All seven business threads have been transformed into automated test scripts. I also ran baseline and 10 concurrent virtual user tests. Each virtual user emulates the exact HTTP requests a real user would make if following the steps outlined in the test cases. Both tests ran free of errors and response times were minimal. I scheduled a 20 concurrent user test via LoadRunner’s Controller module to begin running at 8:00pm for a 60-minute duration. The results from this test will be gathered and analyzed tomorrow morning.
9:30 p.m.
I arrive at Woodhaven Lanes, the bowling alley where I have been going since my senior year in high school. I have played on a league there every Wednesday night since then. This league is a competitive, mixed-gender, money league. So not only will I have lots of fun, but I might make some extra cash. It turns out that it hasn’t been one of my best nights bowling, but I still enjoyed myself.
Thursday 9:00 a.m. (Client Site)
I gather and analyze the results from the 20 concurrent user test that ran last night. Various application errors occurred due to bad location data. Unfortunately, the results had to be discarded. I re-execute the 20-user test since the first test was undermined by bad location data, first ensuring that the data in the system this time was clean, and accurate according to specifications. The test was load balanced and results showed that there were no application errors and response times were optimal.
1:00 p.m.
I run a 50 concurrent user test. There were various application errors due to database locking and a shortage of WebLogic (the underlying application server) threads. An exponential increase in transaction response time was also observed. I choose to re-run the 50-user test to confirm previous encountered issues. Similar database and thread issues were generated for a second time. I spoke with the middleware contact concerning the results. He will be increasing the WebLogic threads and database connections. There will be a meeting later this afternoon to discuss the results of the tests conducted so far, and determine what additional tests we should run. After the meeting I am going to do some more load testing.
5:00 p.m.
I attend the status briefing to discuss the load testing results and moving forward. We decide to adjust one of the test scripts to remove one of the reports needed by the payroll team. This particular report is resource intensive and the applications Subject Matter Experts feel that is highly unlikely it ever be run during peak operation times when end-users are calculating employee payment information. Users have been discouraged from running the report unless an extreme situation arises and it is a necessity. After the meeting, I run a 20 concurrent user load test, excluding the script for this report. After analyzing the new results, the test ran error-free and response times were minimal. To wrap up the day, I scheduled a 50 concurrent user test to run during the night. These results will be analyzed tomorrow morning, and assuming the test executes successfully, a final 100 concurrent user test will run to finalize the second phase of the testing process.
8:30 p.m.
I arrive at the “Y” in Astoria, Queens. I am a member of a basketball league that plays there every week. Tonight’s game is against the first place team in our league. They have a lot of “size” on us, but hopefully our speed and defense will prevail. I have my fingers crossed.
Friday 9:00 a.m. (Client Site)
Driving to the client site this morning, I feel a bit sore from last night’s game. We really put our all into it and came out ahead by 2 points, feeling triumphant over the league’s best. The 50 concurrent user test that ran last night ran successfully and met the time requirements. I ran a 100 concurrent user test, carefully tweaking virtual user ramp-up time to avoid overloading the system. Analysis of the tests results is based off of statistics gathered from the window that exists between when ramp-up completes and before ramp-down begins. Implementing this method ensures that the metrics I obtain are included only from when the system is fully saturated with the desired user load.
10:30 a.m.
The final 100 concurrent user test ran successfully. The test ran error-free and response times were acceptable, even though they have noticeably increased. Doing a final analysis of all test results revealed that the increase in response times was in fact linear. I was able to base my conclusions on a comparison of the response times and virtual users’ graphs, as well as various other graphing methods provided by the LoadRunner Analysis Module.
11:15 a.m.
I began working on my final report for this project. The report includes a background of the project, the testing strategy implemented, my observations and findings, and recommendations for further testing.
5:30 p.m.
My report is complete and I send it to my manager for approval. My purpose at this client site was to provide value and insight concerning this system under test. Even though the application has passed testing for the Canadian release, the release of application in the United States will add an additional 350 users to the population, a load level of which the application cannot currently support. I suggested that optimization should be geared toward fine-tuning the SQL on the backend of one of the data-intensive business transactions and that the developers should look further into the LDAP/Authentication bottleneck that was reoccurring.
6:15 p.m.
I walk to the visitor parking lot where my car is located to begin my commute home. As I am about halfway to my car, I encounter a pack of wild turkeys, which in my opinion, were about the size of ostriches. As we stare each other down, I contemplate whether my brain is playing tricks on me after a long, excitement-filled week. No, they are definitely real I conclude, as an exiting car approaches and the very intimidating pack of wild turkeys skedaddles into the woods. I am a city-boy after all.
7:00 p.m.
Home sweet home
The project is complete.
Another business week comes to a close and the weekend is here.