ExternalVendorCode/openmct
1
0
Fork 0
mirror of https://github.com/nasa/openmct.git synced 2025-06-06 09:21:43 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Merge pull request #349 from nasa/open1573

Browse Source 
[Documentation] Document test process
This commit is contained in:
akhenry 2015-12-04 14:34:59 -08:00
commit dd0f9ab74e
5 changed files with 469 additions and 155 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
CONTRIBUTING.md
View File

@ -291,7 +291,7 @@ checklist.)
1. Changes address original issue? 1. Changes address original issue?
2. Unit tests included and/or updated with changes? 2. Unit tests included and/or updated with changes?
3. Command line build passes? 3. Command line build passes?
4. Expect to pass code review? 4. Changes have been smoke-tested?
### Reviewer Checklist ### Reviewer Checklist

161
docs/src/process/cycle.md Normal file
View File

@ -0,0 +1,161 @@
# Development Cycle
Development of Open MCT Web occurs on an iterative cycle of
sprints and releases.
* A _sprint_ is three weeks in duration, and represents a
set of improvements that can be completed and tested by the
development team. Software at the end of the sprint is
"semi-stable"; it will have undergone reduced testing and may carry
defects or usability issues of lower severity, particularly if
there are workarounds.
* A _release_ occurs every four sprints. Releases are stable, and
will have undergone full acceptance testing to ensure that the
software behaves correctly and usably.
## Roles
The sprint process assumes the presence of a __project manager.__
The project manager is responsible for
making tactical decisions about what development work will be
performed, and for coordinating with stakeholders to arrive at
higher-level strategic decisions about desired functionality
and characteristics of the software, major external milestones,
and so forth.
In the absence of a dedicated project manager, this role may be rotated
among members of the development team on a per-sprint basis.
Responsibilities of the project manager including:
* Maintaining (with agreement of stakeholders) a "road map" of work
planned for future releases/sprints; this should be higher-level,
usually expressed as "themes",
with just enough specificity to gauge feasibility of plans,
relate work back to milestones, and identify longer-term
dependencies.
* Determining (with assistance from the rest of the team) which
issues to work on in a given sprint and how they shall be
assigned.
* Pre-planning subsequent sprints to ensure that all members of the
team always have a clear direction.
* Scheduling and/or ensuring adherence to
[process points](#process-points).
* Responding to changes within the sprint (shifting priorities,
new issues) and re-allocating work for the sprint as needed.
## Sprint Calendar
Certain [process points](#process-points) are regularly scheduled in
the sprint cycle.
### Sprints by Release
Allocation of work among sprints should be planned relative to release
goals and milestones. As a general guideline, higher-risk work (large
new features which may carry new defects, major refactoring, design
changes with uncertain effects on usability) should be allocated to
earlier sprints, allowing for time in later sprints to ensure stability.
| Sprint | Focus |
|:------:|:--------------------------------------------------------|
| __1__ | Prototyping, design, experimentation. |
| __2__ | New features, refinements, enhancements. |
| __3__ | Feature completion, low-risk enhancements, bug fixing. |
| __4__ | Stability & quality assurance. |
### Sprints 1-3
The first three sprints of a release are primarily centered around
development work, with regular acceptance testing in the third
week. During this third week, the top priority should be passing
acceptance testing (e.g. by resolving any blockers found); any
resources not needed for this effort should be used to begin work
for the subsequent sprint.
| Week | Mon | Tue | Wed | Thu | Fri |
|:-----:|:-------------------------:|:------:|:---:|:----------------------------:|:-----------:|
| __1__ | Sprint plan | Tag-up | | | |
| __2__ | | Tag-up | | | Code freeze |
| __3__ | Per-sprint testing | Triage | | _Per-sprint testing*_ | Ship |
* If necessary.
### Sprint 4
The software must be stable at the end of the fourth sprint; because of
this, the fourth sprint is scheduled differently, with a heightened
emphasis on testing.
| Week | Mon | Tue | Wed | Thu | Fri |
|-------:|:-------------------------:|:------:|:---:|:----------------------------:|:-----------:|
| __1__ | Sprint plan | Tag-up | | | Code freeze |
| __2__ | Per-release testing | Triage | | | |
| __3__ | _Per-release testing*_ | Triage | | _Per-release testing*_ | Ship |
* If necessary.
## Process Points
* __Sprint plan.__ Project manager allocates issues based on
theme(s) for sprint, then reviews with team. Each team member
should have roughly two weeks of work allocated (to allow time
in the third week for testing of work completed.)
* Project manager should also sketch out subsequent sprint so
that team may begin work for that sprint during the
third week, since testing and blocker resolution is unlikely
to require all available resources.
* __Tag-up.__ Check in and status update among development team.
May amend plan for sprint as-needed.
* __Code freeze.__ Any new work from this sprint
(features, bug fixes, enhancements) must be integrated by the
end of the second week of the sprint. After code freeze
(and until the end of the sprint) the only changes that should be
merged into the master branch should directly address issues
needed to pass acceptance testing.
* [__Per-release Testing.__](testing/plan.md#per-release-testing)
Structured testing with predefined
success criteria. No release should ship without passing
acceptance tests. Time is allocated in each sprint for subsequent
rounds of acceptance testing if issues are identified during a
prior round. Specific details of acceptance testing need to be
agreed-upon with relevant stakeholders and delivery recipients,
and should be flexible enough to allow changes to plans
(e.g. deferring delivery of some feature in order to ensure
stability of other features.) Baseline testing includes:
* [__Testathon.__](testing/plan.md#user-testing)
Multi-user testing, involving as many users as
is feasible, plus development team. Open-ended; should verify
completed work from this sprint, test exploratorily for
regressions, et cetera.
* [__Long-Duration Test.__](testing/plan.md#long-duration-testing) A
test to verify that the software remains
stable after running for longer durations. May include some
combination of automated testing and user verification (e.g.
checking to verify that software remains subjectively
responsive at conclusion of test.)
* [__Unit Testing.__](testing/plan.md#unit-testing)
Automated testing integrated into the
build. (These tests are verified to pass more often than once
per sprint, as they run before any merge to master, but still
play an important role in per-release testing.)
* [__Per-sprint Testing.__](testing/plan.md#per-sprint-testing)
Subset of Pre-release Testing
which should be performed before shipping at the end of any
sprint. Time is allocated for a second round of
Pre-release Testing if the first round is not passed.
* __Triage.__ Team reviews issues from acceptance testing and uses
success criteria to determine whether or not they should block
release, then formulates a plan to address these issues before
the next round of acceptance testing. Focus here should be on
ensuring software passes that testing in order to ship on time;
may prefer to disable malfunctioning components and fix them
in a subsequent sprint, for example.
* __Ship.__ Tag a code snapshot that has passed acceptance
testing and deploy that version. (Only true if acceptance
testing has passed by this point; if acceptance testing has not
been passed, will need to make ad hoc decisions with stakeholders,
e.g. "extend the sprint" or "defer shipment until end of next
sprint.")

165
docs/src/process/index.md
View File

@ -1,156 +1,13 @@
# Development Cycle # Development Process
Development of Open MCT Web occurs on an iterative cycle of
sprints and releases.
* A _sprint_ is three weeks in duration, and represents a
set of improvements that can be completed and tested by the
development team. Software at the end of the sprint is
"semi-stable"; it will have undergone reduced testing and may carry
defects or usability issues of lower severity, particularly if
there are workarounds.
* A _release_ occurs every four sprints. Releases are stable, and
will have undergone full acceptance testing to ensure that the
software behaves correctly and usably.
## Roles
The sprint process assumes the presence of a __project manager.__
The project manager is responsible for
making tactical decisions about what development work will be
performed, and for coordinating with stakeholders to arrive at
higher-level strategic decisions about desired functionality
and characteristics of the software, major external milestones,
and so forth.
In the absence of a dedicated project manager, this role may be rotated
among members of the development team on a per-sprint basis.
Responsibilities of the project manager including:
* Maintaining (with agreement of stakeholders) a "road map" of work
planned for future releases/sprints; this should be higher-level,
usually expressed as "themes",
with just enough specificity to gauge feasibility of plans,
relate work back to milestones, and identify longer-term
dependencies.
* Determining (with assistance from the rest of the team) which
issues to work on in a given sprint and how they shall be
assigned.
* Pre-planning subsequent sprints to ensure that all members of the
team always have a clear direction.
* Scheduling and/or ensuring adherence to
[process points](#process-points).
* Responding to changes within the sprint (shifting priorities,
new issues) and re-allocating work for the sprint as needed.
## Sprint Calendar
Certain [process points](#process-points) are regularly scheduled in
the sprint cycle.
### Sprints by Release
Allocation of work among sprints should be planned relative to release
goals and milestones. As a general guideline, higher-risk work (large
new features which may carry new defects, major refactoring, design
changes with uncertain effects on usability) should be allocated to
earlier sprints, allowing for time in later sprints to ensure stability.
| Sprint | Focus |
|:------:|:--------------------------------------------------------|
| __1__ | Prototyping, design, experimentation. |
| __2__ | New features, refinements, enhancements. |
| __3__ | Feature completion, low-risk enhancements, bug fixing. |
| __4__ | Stability & quality assurance. |
### Sprints 1-3
The first three sprints of a release are primarily centered around
development work, with regular acceptance testing in the third
week. During this third week, the top priority should be passing
acceptance testing (e.g. by resolving any blockers found); any
resources not needed for this effort should be used to begin work
for the subsequent sprint.
| Week | Mon | Tue | Wed | Thu | Fri |
|:-----:|:-------------------------:|:------:|:---:|:----------------------------:|:-----------:|
| __1__ | Sprint plan | Tag-up | | | |
| __2__ | | Tag-up | | | Code freeze |
| __3__ | Sprint acceptance testing | Triage | | _Sprint acceptance testing*_ | Ship |
* If necessary.
### Sprint 4
The software must be stable at the end of the fourth sprint; because of
this, the fourth sprint is scheduled differently, with a heightened
emphasis on testing.
| Week | Mon | Tue | Wed | Thu | Fri |
|-------:|:-------------------------:|:------:|:---:|:----------------------------:|:-----------:|
| __1__ | Sprint plan | Tag-up | | | Code freeze |
| __2__ | Acceptance testing | Triage | | | |
| __3__ | _Acceptance testing*_ | Triage | | _Acceptance testing*_ | Ship |
* If necessary.
## Process Points
* __Sprint plan.__ Project manager allocates issues based on
theme(s) for sprint, then reviews with team. Each team member
should have roughly two weeks of work allocated (to allow time
in the third week for testing of work completed.)
* Project manager should also sketch out subsequent sprint so
that team may begin work for that sprint during the
third week, since testing and blocker resolution is unlikely
to require all available resources.
* __Tag-up.__ Check in and status update among development team.
May amend plan for sprint as-needed.
* __Code freeze.__ Any new work from this sprint
(features, bug fixes, enhancements) must be integrated by the
end of the second week of the sprint. After code freeze
(and until the end of the sprint) the only changes that should be
merged into the master branch should directly address issues
needed to pass acceptance testing.
* __Acceptance Testing.__ Structured testing with predefined
success criteria. No release should ship without passing
acceptance tests. Time is allocated in each sprint for subsequent
rounds of acceptance testing if issues are identified during a
prior round. Specific details of acceptance testing need to be
agreed-upon with relevant stakeholders and delivery recipients,
and should be flexible enough to allow changes to plans
(e.g. deferring delivery of some feature in order to ensure
stability of other features.) Baseline testing includes:
* __Testathon.__ Multi-user testing, involving as many users as
is feasible, plus development team. Open-ended; should verify
completed work from this sprint, test exploratorily for
regressions, et cetera.
* __24-Hour Test.__ A test to verify that the software remains
stable after running for longer durations. May include some
combination of automated testing and user verification (e.g.
checking to verify that software remains subjectively
responsive at conclusion of test.)
* __Automated Testing.__ Automated testing integrated into the
build. (These tests are verified to pass more often than once
per sprint, as they run before any merge to master, but still
play an important role in acceptance testing.)
* __Sprint Acceptance Testing.__ Subset of Acceptance Testing
which should be performed before shipping at the end of any
sprint. Time is allocated for a second round of
Sprint Acceptance Testing if the first round is not passed.
* __Triage.__ Team reviews issues from acceptance testing and uses
success criteria to determine whether or not they should block
release, then formulates a plan to address these issues before
the next round of acceptance testing. Focus here should be on
ensuring software passes that testing in order to ship on time;
may prefer to disable malfunctioning components and fix them
in a subsequent sprint, for example.
* __Ship.__ Tag a code snapshot that has passed acceptance
testing and deploy that version. (Only true if acceptance
testing has passed by this point; if acceptance testing has not
been passed, will need to make ad hoc decisions with stakeholders,
e.g. "extend the sprint" or "defer shipment until end of next
sprint.")
The process used to develop Open MCT Web is described in the following
documents:
* [Development Cycle](cycle.md): Describes how and when specific
process points are repeated during development.
* Testing is described in two documents:
* The [Test Plan](testing/plan.md) summarizes the approaches used
to test Open MCT Web.
* The [Test Procedures](testing/procedures.md) document what
specific tests are performed to verify correctness, and how
they should be carried out.

127
docs/src/process/testing/plan.md Normal file
View File

@ -0,0 +1,127 @@
# Test Plan
## Test Levels
Testing for Open MCT Web includes:
* _Smoke testing_: Brief, informal testing to verify that no major issues
or regressions are present in the software, or in specific features of
the software.
* _Unit testing_: Automated verification of the performance of individual
software components.
* _User testing_: Testing with a representative user base to verify
that application behaves usably and as specified.
* _Long-duration testing_: Testing which takes place over a long period
of time to detect issues which are not readily noticeable during
shorter test periods.
### Smoke Testing
Manual, non-rigorous testing of the software and/or specific features
of interest. Verifies that the software runs and that basic functionality
is present.
### Unit Testing
Unit tests are automated tests which exercise individual software
components. Tests are subject to code review along with the actual
implementation, to ensure that tests are applicable and useful.
Unit tests should meet
[test standards](https://github.com/nasa/openmctweb/blob/master/CONTRIBUTING.md#test-standards)
as described in the contributing guide.
### User Testing
User testing is performed at scheduled times involving target users
of the software or reasonable representatives, along with members of
the development team exercising known use cases. Users test the
software directly; the software should be configured as similarly to
its planned production configuration as is feasible without introducing
other risks (e.g. damage to data in a production instance.)
User testing will focus on the following activities:
* Verifying issues resolved since the last test session.
* Checking for regressions in areas related to recent changes.
* Using major or important features of the software,
as determined by the user.
* General "trying to break things."
During user testing, users will
[report issues](https://github.com/nasa/openmctweb/blob/master/CONTRIBUTING.md#issue-reporting)
as they are encountered.
Desired outcomes of user testing are:
* Identified software defects.
* Areas for usability improvement.
* Feature requests (particularly missed requirements.)
* Recorded issue verification.
### Long-duration Testing
Long-duration testing occurs over a twenty-four hour period. The
software is run in one or more stressing cases representative of expected
usage. After twenty-four hours, the software is evaluated for:
* Performance metrics: Have memory usage or CPU utilization increased
during this time period in unexpected or undesirable ways?
* Subjective usability: Does the software behave in the same way it did
at the start of the test? Is it as responsive?
Any defects or unexpected behavior identified during testing should be
[reported as issues](https://github.com/nasa/openmctweb/blob/master/CONTRIBUTING.md#issue-reporting)
and reviewed for severity.
## Test Performance
Tests are performed at various levels of frequency.
* _Per-merge_: Performed before any new changes are integrated into
the software.
* _Per-sprint_: Performed at the end of every [sprint](../cycle.md).
* _Per-release_: Performed at the end of every [release](../cycle.md).
### Per-merge Testing
Before changes are merged, the author of the changes must perform:
* _Smoke testing_ (both generally, and for areas which interact with
the new changes.)
* _Unit testing_ (as part of the automated build step.)
Changes are not merged until the author has affirmed that both
forms of testing have been performed successfully; this is documented
by the [Author Checklist](https://github.com/nasa/openmctweb/blob/master/CONTRIBUTING.md#author-checklist).
### Per-sprint Testing
Before a sprint is closed, the development team must additionally
perform:
* A relevant subset of [_user testing_](procedures.md#user-test-procedures)
identified by the acting [project manager](../cycle.md#roles).
* [_Long-duration testing_](procedures.md#long-duration-testng)
(specifically, for 24 hours.)
Issues are reported as a product of both forms of testing.
A sprint is not closed until both categories have been performed on
the latest snapshot of the software, _and_ no issues labelled as
["blocker"](https://github.com/nasa/openmctweb/blob/master/CONTRIBUTING.md#issue-reporting)
remain open.
### Per-release Testing
As [per-sprint testing](#per-sprint-testing), except that _user testing_
should cover all test cases, with less focus on changes from the specific
sprint or release.
Per-release testing should also include any acceptance testing steps
agreed upon with recipients of the software.
A release is not closed until both categories have been performed on
the latest snapshot of the software, _and_ no issues labelled as
["blocker" or "critical"](https://github.com/nasa/openmctweb/blob/master/CONTRIBUTING.md#issue-reporting)
remain open.

169
docs/src/process/testing/procedures.md Normal file
View File

@ -0,0 +1,169 @@
# Test Procedures
## Introduction
This document is intended to be used:
* By testers, to verify that Open MCT Web behaves as specified.
* By the development team, to document new test cases and to provide
guidance on how to author these.
## Writing Procedures
### Template
Procedures for individual tests should use the following template,
adapted from [https://swehb.nasa.gov/display/7150/SWE-114]().
Property | Value
---------------|---------------------------------------------------------------
Test ID |
Relevant reqs. |
Prerequisites |
Test input |
Instructions |
Expectation |
Eval. criteria |
For multi-line descriptions, use an asterisk or similar indicator to refer
to a longer-form description below.
#### Example Procedure - Edit a Layout
Property | Value
---------------|---------------------------------------------------------------
Test ID | MCT-TEST-000X - Edit a layout
Relevant reqs. | MCT-EDIT-000Y
Prerequisites | Create a layout, as in MCT-TEST-000Z
Test input | Domain object database XYZ
Instructions | See below *
Expectation | Change to editing context †
Eval. criteria | Visual inspection
* Follow the following steps:
1. Verify that the created layout is currently navigated-to,
as in MCT-TEST-00ZZ.
2. Click the Edit button, identified by a pencil icon and the text "Edit"
displayed on hover.
† Right-hand viewing area should be surrounded by a dashed
blue border when a domain object is being edited.
### Guidelines
Test procedures should be written assuming minimal prior knowledge of the
application: Non-standard terms should only be used when they are documented
in [the glossary](#glossary), and shorthands used for user actions should
be accompanied by useful references to test procedures describing those
actions (when available) or descriptions in user documentation.
Test cases should be narrow in scope; if a list of steps is excessively
long (or must be written vaguely to be kept short) it should be broken
down into multiple tests which reference one another.
All requirements satisfied by Open MCT Web should be verifiable using
one or more test procedures.
## Glossary
This section will contain terms used in test procedures. This may link to
a common glossary, to avoid replication of content.
## Procedures
This section will contain specific test procedures. Presently, procedures
are placeholders describing general patterns for setting up and conducting
testing.
### User Testing Setup
These procedures describes a general pattern for setting up for user
testing. Specific deployments should customize this pattern with
relevant data and any additional steps necessary.
Property | Value
---------------|---------------------------------------------------------------
Test ID | MCT-TEST-SETUP0 - User Testing Setup
Relevant reqs. | TBD
Prerequisites | Build of relevant components
Test input | Exemplary database; exemplary telemetry data set
Instructions | See below
Expectation | Able to load application in a web browser (Google Chrome)
Eval. criteria | Visual inspection
Instructions:
1. Start telemetry server.
2. Start ElasticSearch.
3. Restore database snapshot to ElasticSearch.
4. Start telemetry playback.
5. Start HTTP server for client sources.
### User Test Procedures
Specific user test cases have not yet been authored. In their absence,
user testing is conducted by:
* Reviewing the text of issues from the issue tracker to understand the
desired behavior, and exercising this behavior in the running application.
(For instance, by following steps to reproduce from the original issue.)
* Issues which appear to be resolved should be marked as such with comments
on the original issue (e.g. "verified during user testing MM/DD/YYYY".)
* Issues which appear not to have been resolved should be reopened with an
explanation of what unexpected behavior has been observed.
* In cases where an issue appears resolved as-worded but other related
undesirable behavior is observed during testing, a new issue should be
opened, and linked to from a comment in the original issues.
* General usage of new features and/or existing features which have undergone
recent changes. Defects or problems with usability should be documented
by filing issues in the issue tracker.
* Open-ended testing to discover defects, identify usability issues, and
generate feature requests.
### Long-Duration Testing
The purpose of long-duration testing is to identify performance issues
and/or other defects which are sensitive to the amount of time the
application is kept running. (Memory leaks, for instance.)
Property | Value
---------------|---------------------------------------------------------------
Test ID | MCT-TEST-LDT0 - Long-duration Testing
Relevant reqs. | TBD
Prerequisites | MCT-TEST-SETUP0
Test input | (As for test setup.)
Instructions | See "Instructions" below *
Expectation | See "Expectations" below †
Eval. criteria | Visual inspection
* Instructions:
1. Start `top` or a similar tool to measure CPU usage and memory utilization.
2. Open several user-created displays (as many as would be realistically
opened during actual usage in a stressing case) in some combination of
separate tabs and windows (approximately as many tabs-per-window as
total windows.)
3. Ensure that playback data is set to run continuously for at least 24 hours
(e.g. on a loop.)
4. Record CPU usage and memory utilization.
5. In at least one tab, try some general user interface gestures and make
notes about the subjective experience of using the application. (Particularly,
the degree of responsiveness.)
6. Leave client displays open for 24 hours.
7. Record CPU usage and memory utilization again.
8. Make additional notes about the subjective experience of using the
application (again, particularly responsiveness.)
9. Check logs for any unexpected warnings or errors.
† Expectations:
* At the end of the test, CPU usage and memory usage should both be similar
to their levels at the start of the test.
* At the end of the test, subjective usage of the application should not
be observably different from the way it was at the start of the test.
(In particular, responsiveness should not decrease.)
* Logs should not contain any unexpected warnings or errors ("expected"
warnings or errors are those that have been documented and prioritized
as known issues, or those that are explained by transient conditions
external to the software, such as network outages.)