CORTX-TEST is a repository for multiple automation projects developed for testing CORTX and supported solutions/systems. These frameworks are reusable for any opensource object stores with minimal configurations.
It is logically divided into following components:
- Test Automation framework (Corbot)
- Test Execution Framework (drunner)
- Robot framework UI Tests and
- Tools ( Reporting Dashboards, Data Integrity, Clone Test Plans etc.)
Architectural Overview page describes the architectural considerations and repository layout. Test Execution Deployment View describes the deployment design view of the test framework in distributed mode. Test Framework Design Document describes in details the design of the framework in distributed mode.
This document assumes that you are aware about Github and if you are coming from SVN or other code versioning system it is recommended to follow the link Github Process Readme to configure git on your local machine. Following Readme document will give you enough insights to start contributing.
You need a separate client VM with any Linux flavour (prefer CentOS 7+ ) to install client side pre-requisites and start running automation framework on the same VM. This VM should have network connectivity to a Cortx Cluster OR CORTX OVA deployment. Alternatively you may use one of the nodes as client (less recommended).
Typically, a member contributing to test framework would follow the review process as follows:
- Commits happen on your forked repository(origin).
- Then you can raise a PR to merge it to Seagate Cortx-Test repository (Upstream). PRs can be raised even if you have read-only access to Seagate Repositories.
- Moderators of Cortx-Test can create server side feature branch if multiple developers are working on same feature branch
- Team member can check-in and raise the PR to upstream feature branch. Feature lead would raise the pull request to main.
Fork local repository from Seagate's Cortx-Test repository and then clone Cortx-Test repository from Seagate repository. Commands as follows:
git clone https://github.com/Seagate/cortx-test.git
cd cortx-test/
git status
git branch
git checkout main
git remote -v
git remote add upstream https://github.com/Seagate/cortx-test.git
git remote -v
Issuing the above command again will return output as shown:
> origin https://github.com/YOUR_USERNAME/cortx-test.git (fetch)
> origin https://github.com/YOUR_USERNAME/cortx-test.git (push)
> upstream https://github.com/Seagate/cortx-test.git (fetch)
> upstream https://github.com/Seagate/cortx-test.git (push)
Then fetch upstream...
git fetch upstream
git pull upstream main
Following steps help in setting up client side env, where test framework will run. These steps assume that you have followed earlier process (i.e. git client is installed and Cortx-test is cloned)
1. `yum update -y`
2. `yum install -y gcc openssl-devel bzip2-devel libffi-devel zlib-devel wget make sqlite-devel`
3. `cd /usr/src && wget https://www.python.org/ftp/python/3.7.9/Python-3.7.9.tgz && tar xzf Python-3.7.9.tgz && rm Python-3.7.9.tgz`
4. `cd /usr/src/Python-3.7.9 && ./configure --prefix=/usr --enable-optimizations`
4a. Note if you get an error message:
configure: error: in `/usr/src/Python-3.7.9':
configure: error: no acceptable C compiler found in $PATH
See `config.log' for more details
You need to install gcc:
- Redhat base:
`yum groupinstall "Development Tools"`
- Debian base:
`apt-get install build-essential`
- openSUSE base:
`zypper install --type pattern devel_basis`
5. `cd /usr/src/Python-3.7.9 && make altinstall`
5a. Right here at this point you can check python is installed correctly by going in interactive mode. You can issue command "pip3 install pysqlite3" and type "import sqlite3" to confirm that sqlite3 is installed. This will save a lot of your time if you run into issues later for python installation. In some linux flavours you need --enable-loadable-sqlite-extensions switch to be added while configuring python.
cd /usr/src/Python-3.7.9 && ./configure --prefix=/usr --enable-optimizations --enable-loadable-sqlite-extensions
6. Create a softlink to point to this installation. You can check the folder created and
improvise following command.
`ln -s /usr/local/bin/python3.7 python3.7
7. `yum install -y python3-devel librdkafka nfs-utils python3-tkinter`
8. `python3.7 -m venv virenv`
9. `source virenv/bin/activate` or use old style `. ./virenv/bin/activate`
8. `pip install --upgrade pip`
9. `pip install pysqlite3`
10. Change dir to cortx-test project directory, make sure a requirement file is present in project dir. Use following command to install python packages.
`pip install --ignore-installed -r requirements.txt`
You can issue virenv/bin/deactivate to deactivate pyenv.
11. Install awscli with default python 3.6 pre installed with inhouse vm images and
configure aws and copy cert file.
Alternatively by skipping step 8 to 10, you can also set python environment by using virtual env.
Change dir to your local repository root folder. If you have checked out your code
in clean_dev directory created in your home on Linux machine (RHEL Flavour), then
/home/<yourname>/clean_dev is the local repository root folder.
# cd clean_dev
# ./cortx-test/ci_tools/client_setup.sh
This script should handle client setup. However, note that python configures does not have switch --enable-loadable-sqlite-extensions in script.
mkdir -p /etc/ssl/stx
mkdir -p /etc/ssl/stx-s3-clients/s3/
curl https://raw.githubusercontent.com/Seagate/cortx-s3server/kubernetes/scripts/haproxy/ssl/s3.seagate.com.crt -o /etc/ssl/stx-s3-clients/s3/ca.crt
curl https://raw.githubusercontent.com/Seagate/cortx-prvsnr/4c2afe1c19e269ecb6fbf1cba62fdb7613508182/srv/components/misc_pkgs/ssl_certs/files/stx.pem -o /etc/ssl/stx/stx.pem
To set up s3 client tools, make sure you have completed basic setup in Set up dev environment.
Script in project's root folder cortx-test scripts/s3_tools/Makefile can be used to install s3 tools on client.
Required arguments in configuration:
ACCESS aws_access_key_id
SECRET aws_secret_access_key
optional arguments:
-i --ignore-errors Ignore all errors in commands executed to remake files.
-k --keep-going Continue as much as possible after an error.
--ENDPOINT=s3_endpoint
--CA_CRT=certificate_file_path
--NFS_SHARE=NFS_share_jclient_path
--APACHE_J_METER=apache-jmeter-5.4.1.tgz
--VERIFY_SSL='True' This is used to whether https/ssl be used or not
--VALIDATE_CERTS='True' This is used whether given certificate should be verified or not.
make help --makefile=scripts/s3_tools/Makefile
all : Install & configure tools like aws, s3fs, s3cmd, minio, call in case it's a new machine. Eg: make all ACCESS=<new-accesskey> SECRET=<new-secretkey>
clean : Remove installed tools like aws, s3fs, s3cmd, minio. Eg: make clean
install-tools : Install tools like aws, s3fs, s3cmd, minio, call in case it's a new machine. Eg: make install-tools
configure-tools: Install tools like aws, s3fs, s3cmd, minio, call in case it's a new machine. Eg: make configure-tools ACCESS=<new-accesskey> SECRET=<new-secretkey>
aws : Install & configure aws tool. Eg: make aws ACCESS=<new-accesskey> SECRET=<new-secretkey>
s3fs : Install & configure s3fs tool. Eg: make s3fs ACCESS=<new-accesskey> SECRET=<new-secretkey>
s3cmd : Install & configure s3cmd tool. Eg: make s3cmd ACCESS=<new-accesskey> SECRET=<new-secretkey>
jcloud-client: Setup jcloud-client. Eg: make jcloud-client
minio : Install & configure minio tool. Eg: make minio ACCESS=<new-accesskey> SECRET=<new-secretkey>
s3bench-install: Setup s3bench tool. Eg: make s3bench-install --makefile=<makefile_path>"
apache-jmeter-install: Setup apache-jmeter-install tool. Eg: make apache-jmeter-install --makefile=<makefile_path>"
bashrc-configure : Configure ~/.bashrc for updating ulimit -n for allowing file descriptors
To increase ulimit for allowing maximum file descriptors
make bashrc-configure
To install & configure all tools:
make all --makefile=scripts/s3_tools/Makefile ACCESS=<aws_access_key_id> SECRET=<aws_secret_access_key> ENDPOINT=<lb ipaddress> VALIDATE_CERTS=<Tree/False> VERIFY_SSL=<True/False>
To just configure installed tools:
make configure-tools --makefile=scripts/s3_tools/Makefile ACCESS=<aws_access_key_id> SECRET=<aws_secret_access_key> ENDPOINT=<lb ipaddress> VALIDATE_CERTS=<Tree/False> VERIFY_SSL=<True/False>
To install & configure specific tool(i.e aws):
make aws --makefile=scripts/s3_tools/Makefile ACCESS=<aws_access_key_id> SECRET=<aws_secret_access_key> ENDPOINT=<lb ipaddress> VALIDATE_CERTS=<Tree/False> VERIFY_SSL=<True/False>
To just configure specific tool(i.e. aws):
make aws-configure --makefile=scripts/s3_tools/Makefile ACCESS=<aws_access_key_id> SECRET=<aws_secret_access_key> ENDPOINT=<lb ipaddress> VALIDATE_CERTS=<Tree/False> VERIFY_SSL=<True/False>
To clean up all tools:
make clean --makefile=scripts/s3_tools/Makefile
Cortx-test uses MongoDB as backend to store Cortx setup details. These details, stored in MongoDB, are specific
to the setup itself. The purpose of this setup is to do automatic config generation
based on the setup. Not all values are mandatory and only applicable values needs to be filled in vm environment. A sample template is as shown below. This template is fed to the database and pulled when developer will run test automation with test runner. The pulled templates merge with static yaml files to build the CMN_CFG and other component level configs.
If you don't want to store or use MongoDB configuration entry, you may create setups.json locally in project root cortx-test whch will be used to build and load configs in test run process.
While testing with cortx-test in dev environment or Community you may like to have setups.json created and skip MongoDB interaction.
Setups.json format is shown below. Sample MongoDB entry under A sample MongoDB entry for a target setup. is placed in value part of LC_Setup dict shown below.
{"LC_Setup": {
}
}A sample MongoDB entry for a target setup.
{
"setupname":"LC_Setup",
"setup_type":"VM",
"setup_in_useby": "",
"in_use_for_parallel": false,
"parallel_client_cnt": 0,
"is_setup_free": true,
"product_family": "LC",
"s3_engine": 1,
"product_type": "k8s",
"lb": "FQDN without protocol(http/s)",
"nodes":[
{
"host": "srv-node-0",
"hostname": "node 0 hostname",
"ip": "node 0 ip",
"fqdn": "node 0 fqdn",
"username": "node 0 username",
"password": "node 0 password",
"public_data_ip":"",
"private_data_ip":"",
"node_type":"master",
"lpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"rpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"encl_lpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"encl_rpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"gem_controller": {
"ip": "",
"user": "",
"pwd": "",
"port1": "",
"port2": ""
}
},
{
"host": "srv-node-1",
"hostname": "node 1 hostname",
"ip": "node 1 ip address",
"fqdn": "node 0 fqdn",
"username": "node 1 username",
"password": "node 1 password",
"public_data_ip":"",
"private_data_ip":"",
"node_type":"worker",
"lpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"rpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"encl_lpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"encl_rpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"gem_controller": {
"ip": "",
"user": "",
"pwd": "",
"port1": "",
"port2": ""
}
},
{
"host": "srv-node-2",
"hostname": "node 2 hostname",
"ip": "node 2 ip address",
"fqdn": "node 0 fqdn",
"username": "node 2 username",
"password": "node 2 password",
"public_data_ip":"",
"private_data_ip":"",
"node_type":"worker",
"lpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"rpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"encl_lpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"encl_rpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"gem_controller": {
"ip": "",
"user": "",
"pwd": "",
"port1": "",
"port2": ""
}
}
],
"csm":
{
"mgmt_vip": "fqdn",
"port": "31169",
"csm_admin_user":{
"username": "",
"password": ""
}
},
"bmc":
{
"username": "",
"password": ""
},
"s3endpoints":
{
"s3_io": "fqdn:port",
"s3_auth": "fqdn:port"
}
}Following json shows a minimal example setup configuration. You may want to refer full json at tools/setup_update/setup_entry_lc.sample.json.
{
"setupname":"Unique_name/FQDN",
"setup_type":"VM",
"product_family": "LC",
"product_type": "k8s",
"s3_engine": 2,
"lb": "FQDN of LB or K8S SVC",
"nodes":[
{
"host": "srv-node-0",
"hostname": "FQDN",
"ip": "IPV4",
"fqdn": "FQDN",
"username": "root",
"password": "",
"public_data_ip":"",
"private_data_ip":"",
"node_type":"master",
"lpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"rpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"encl_lpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"encl_rpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"gem_controller": {
"ip": "",
"user": "",
"pwd": "",
"port1": "",
"port2": ""
}
},
{
"host": "srv-node-1",
"hostname": "FQDN",
"ip": "IPV4",
"fqdn": "FQDN",
"username": "root",
"password": "",
"public_data_ip":"",
"private_data_ip":"",
"node_type":"worker",
"lpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"rpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"encl_lpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"encl_rpdu": {
"ip": "",
"port": "",
"user": "",
"pwd": ""
},
"gem_controller": {
"ip": "",
"user": "",
"pwd": "",
"port1": "",
"port2": ""
}
}
],
"csm":
{
"mgmt_vip": "FQDN",
"port": "31169",
"csm_admin_user":{
"username": "<csm_uid>",
"password": ""
}
},
"bmc":
{
"username": "",
"password": ""
}
}
Script in project's path tools/setup_update can be used to generate a setup specific config entry.
python setup_entry.py --help
usage: setup_entry.py [-h] [--fpath FPATH] [--dbuser DBUSER] [--dbpassword DBPASSWORD] [--new_entry NEW_ENTRY]
Update the setup entry
optional arguments:
-h, --help show this help message and exit
--fpath FPATH Path of the json entry file
--dbuser DBUSER Database user
--dbpassword DBPASSWORD
database password
--new_entry NEW_ENTRY
True for new entry , False for update
e.g. python3 tools/setup_update/setup_entry.py --dbuser <> --dbpassword <>
Name of setup specified in json file should be unique in case you are creating a new setup.
For example in sample json setupname value should be unique `"setupname":"T2"`.
If you want to anyway run the parallel tests sequentially. You should use --force_serial_run switch as shown in following command.
python -u testrunner.py -te TEST-17412 -tp TEST-18382 --target s3-vm-2928 --force_serial_run True
If you want to run your test plan and TE ticket with test runner, you should skip --force_serial_run switch.
python -u testrunner.py -te TEST-17412 -tp TEST-18382 --target s3-vm-2928
When you want to run test and don't want to update Report DB or JIRA
python -u testrunner.py -te TEST-17412 -tp TEST-18382 --target s3-vm-2928 --force_serial_run true --db_update False --jira_update False
python -u testrunner.py --help
usage: testrunner.py [-h] [-j JSON_FILE] [-r HTML_REPORT] [-d DB_UPDATE]
[-u JIRA_UPDATE] [-te TE_TICKET] [-pe PARALLEL_EXE]
[-tp TEST_PLAN] [-b BUILD] [-t BUILD_TYPE] [-tg TARGET]
[-ll LOG_LEVEL] [-p PRC_CNT] [-f [FORCE_SERIAL_RUN]]
[-i DATA_INTEGRITY_CHK]
optional arguments:
-h, --help show this help message and exit
-j JSON_FILE, --json_file JSON_FILE
json file name
-r HTML_REPORT, --html_report HTML_REPORT
html report name
-d DB_UPDATE, --db_update DB_UPDATE
Update Reports DB. Can be false in case reports db is
down
-u JIRA_UPDATE, --jira_update JIRA_UPDATE
Update Jira. Can be false in case Jira is down
-te TE_TICKET, --te_ticket TE_TICKET
jira xray test execution id
-pe PARALLEL_EXE, --parallel_exe PARALLEL_EXE
parallel_exe: True for parallel, False for sequential
-tp TEST_PLAN, --test_plan TEST_PLAN
jira xray test plan id
-b BUILD, --build BUILD
Build number
-t BUILD_TYPE, --build_type BUILD_TYPE
Build type (Release/Dev)
-tg TARGET, --target TARGET
Target setup details
-ll LOG_LEVEL, --log_level LOG_LEVEL
log level value
-p PRC_CNT, --prc_cnt PRC_CNT
number of parallel processes
-f [FORCE_SERIAL_RUN], --force_serial_run [FORCE_SERIAL_RUN]
Force sequential run if you face problems with
parallel run
-i DATA_INTEGRITY_CHK, --data_integrity_chk DATA_INTEGRITY_CHK
Helps set DI check enabled so that tests perform
additional checksum check
-pf PRODUCT_FAMILY, --product_family
Helps to select product family type whether LR or local
-c VALIDATE_CERTS, --validate_certs
This gives option whetherValidate HTTPS/SSL certificate
to S3 endpoint needs to be validated or not.
-s USE_SSL, --use_ssl
Option whether HTTPS/SSL connection for S3 endpoint should be used or not.
-csm CSM_CHECKS, --csm_checks
Execute tests with error code & msg check enabled.
-hc HEALTH_CHECK --health_check
Decide whether to do health check (on server) or not with tests execution.
pytest --local=True -d --tx 3*popen -rA unittests\Your_Test_Module.py
3 is number of worker processes to be spawned.pytest --capture=no --te_tkt TEST-17412 -d --tx 2*popen -rA unittests\Your_Test_Module.py
pytest --capture=no --te_tkt TEST-17412 -rA unittests\test_reporting_and_logging_hooks.py
While ordering client on ssc-cloud, make sure
1. Have at least 8GB RAM for it, to support 1GB object size in s3bench tests.
2. For more large number of parallel IO connections, good to have 8 CPUs.
3. Default 1GB of swap space is provided, need to order 1 extra disk, and create swap space out of extra disk, and mount it.
Procedure to create swap space of 8 GB: A gist from one of the articles
* Create new partition using fdisk command
* fdisk /dev/sdb # sda will generally have OS installation
* new (option n), primary (option p) partition, Default partition number, Default first sector, Last Sector +8G, Write (option w)
* Create swap on the partition using mkswap /dev/sdb1 # Provide above created partition number i.e. sdb1
* Mount swap using swapon /dev/sdb1
## Increase client root space size should be at least 50 GB using following commands
Please utilize free disks from the output of lsblk
Note: In case of multipart/Big object upload, disk space requirement may change/increase.
resize2fs is specific to ext2/3/4. In case /root is formatted with xfs we need to use the xfs_growfs tool
e.g. "xfs_growfs /dev/mapper/vg_sysvol-lv_root"
We can run lvextend as "lvextend /dev/mapper/vg_sysvol-lv_root -l +100%FREE" to consume all the free PE
```
df -h
lsblk
pvcreate /dev/sdb
vgextend vg_sysvol /dev/sdb
lvextend /dev/mapper/vg_sysvol-lv_root -L +50G
resize2fs /dev/mapper/vg_sysvol-lv_root
df -h
```
## increase the swap space, Please utilize free disks from the output of lsblk
```
lsblk
pvcreate /dev/sdi
vgextend vg_sysvol /dev/sdi
lvextend /dev/mapper/vg_sysvol-lv_swap -l +100%FREE
swapoff /dev/mapper/vg_sysvol-lv_swap
mkswap /dev/mapper/vg_sysvol-lv_swap
swapon /dev/mapper/vg_sysvol-lv_swap
```