This guide is to introduce how to execute shell_long
test suite via test tool CTP. The shell_long
test case specification is exactly same as general SHELL test case's. The only difference between them is that shell_long
test cases ask a little long time to execute.
Please refer to these sections of shell_heavy guide to execute shell_long test. Below we just list some difference between them.
-
Create shell_long test configuration file as below:
FileCTP/conf/shell_long.conf
# path to a test case folder scenario=${HOME}/cubrid-testcases-private/longcase/shell/1hour/bug_bts_9382
Note: scenario should be a path to a folder but not a file.
-
Check out
shell_long
test cases from git repository in path https://github.com/CUBRID/cubrid-testcases-private/tree/develop/longcase.$ cd ~ $ git clone https://github.com/CUBRID/cubrid-testcases-private.git $ ls ~/cubrid-testcases-private/longcase shell $ cd shell/ $ ls _04_misc _06_issues 1hour 2hour _30_banana_qa _32_features_930 3hour 5hour config other
In this section, we will introduce how to deploy shell_long
regression test environment.
For current daily regression test, there are one controller node and five test nodes.
Controller node : This node listens to test messages and starts a test when there is a test message.
Test node : CUBRID server is deployed on this node, we execute test cases on it.
Information about each test machines.
Description | OS User | IP | Hostname | Tools to deploy |
---|---|---|---|---|
Controller node | rqgcontroller | 192.168.1.99 | func24 | CTP |
Test node | shell_long | 192.168.1.99 | func24 | CTP,cubrid-testcases-private |
Test node | shell_long | 192.168.1.100 | func25 | CTP,cubrid-testcases-private |
Test node | shell_long | 192.168.1.101 | func26 | CTP,cubrid-testcases-private |
Test node | shell_long | 192.168.1.102 | func27 | CTP,cubrid-testcases-private |
Test node | shell_long | 192.168.1.103 | func28 | CTP,cubrid-testcases-private |
-
Please follow the guide to install CTP. Then create shell_long test configuration. File
~/CTP/conf/shell_template_for_shell_long.conf
:$ cat shell_template_for_shell_long.conf default.cubrid.cubrid_port_id=1568 default.broker1.BROKER_PORT=30090 default.broker1.APPL_SERVER_SHM_ID=30090 default.broker2.BROKER_PORT=33091 default.broker2.APPL_SERVER_SHM_ID=33091 default.ha.ha_port_id=59909 env.99.ssh.host=192.168.1.99 env.99.ssh.port=22 env.99.ssh.user=shell_long env.99.ssh.pwd=****** env.100.ssh.host=192.168.1.100 env.100.ssh.port=22 env.100.ssh.user=shell_long env.100.ssh.pwd=****** env.101.ssh.host=192.168.1.101 env.101.ssh.port=22 env.101.ssh.user=shell_long env.101.ssh.pwd=****** env.102.ssh.host=192.168.1.102 env.102.ssh.port=22 env.102.ssh.user=shell_long env.102.ssh.pwd=****** env.103.ssh.host=192.168.1.103 env.103.ssh.port=22 env.103.ssh.user=shell_long env.103.ssh.pwd=****** scenario=${HOME}/cubrid-testcases-private/longcase/shell test_continue_yn=false cubrid_download_url=http://127.0.0.1/REPO_ROOT/store_02/10.1.0.6876-f9026f8/drop/CUBRID-10.1.0.6876-f9026f8-Linux.x86_64.sh testcase_exclude_from_file=${HOME}/cubrid-testcases-private/longcase/shell/config/daily_regression_test_excluded_list_linux.conf testcase_update_yn=true testcase_git_branch=develop #testcase_timeout_in_secs=604800 test_platform=linux test_category=shell_long testcase_exclude_by_macro=LINUX_NOT_SUPPORTED testcase_retry_num=0 delete_testcase_after_each_execution_yn=false enable_check_disk_space_yn=true feedback_type=database feedback_notice_qahome_url=http://192.168.1.86:8080/qaresult/shellImportAction.nhn?main_id=<MAINID> owner_email=<owner e-mail> cc_email=<cc e-mail to receive notice> git_user=<git user> git_email=<git e-mail> git_pwd=****** feedback_db_host=192.168.1.86 feedback_db_port=33080 feedback_db_name=qaresu feedback_db_user=dba feedback_db_pwd=********
Note: when you need to test shell_long_debug, copy
~/CTP/conf/shell_template_for_shell_long.conf
as~/CTP/conf/shell_template_for_shell_long_debug.conf
-
File
~/start_test.sh
# If only need to listen the shell_long test message # nohup start_consumer.sh -q QUEUE_CUBRID_QA_SHELL_LONG_LINUX -exec run_shell & # We use one controllar to listening shell_heavy, shell_long, and RQG test messages in dailyqa. nohup start_consumer.sh -q QUEUE_CUBRID_QA_SHELL_HEAVY_LINUX,QUEUE_CUBRID_QA_RQG,QUEUE_CUBRID_QA_SHELL_LONG_LINUX -exec run_shell,run_shell,run_shell &
-
export DEFAULT_BRANCH_NAME=develop export CTP_HOME=$HOME/CTP export CTP_BRANCH_NAME=develop export CTP_SKIP_UPDATE=0 export init_path=$HOME/CTP/shell/init_path ulimit -c unlimited export LC_ALL=en_US . $HOME/.cubrid.sh export PATH=$CTP_HOME/bin:$CTP_HOME/common/script:$JAVA_HOME/bin:/usr/local/bin:/bin:/usr/bin:$PATH
-
Please follow the guide to install CTP.
-
cd ~ git clone https://github.com/CUBRID/cubrid-testcases-private.git cd ~/cubrid-testcases-private/ git checkout develop
-
$ cat .bash_profile # .bash_profile # Get the aliases and functions if [ -f ~/.bashrc ]; then . ~/.bashrc fi # User specific environment and startup programs PATH=$PATH:$HOME/.local/bin:$HOME/bin export PATH export LC_ALL=en_US export CTP_HOME=$HOME/CTP export LCOV_HOME=/usr/local/cubridqa/lcov-1.11 export init_path=$HOME/CTP/shell/init_path export GCOV_PREFIX=$HOME export GCOV_PREFIX_STRIP=2 export CTP_BRANCH_NAME="develop" export CTP_SKIP_UPDATE="0" export PATH=$CTP_HOME/bin:$CTP_HOME/common/script:$CUBRID/bin:$LCOV_HOME/bin:$JAVA_HOME/bin:$PATH:/usr/local/sbin:/usr/sbin ulimit -c unlimited #------------------------------------------------------------------------------- # set CUBRID environment variables #------------------------------------------------------------------------------- . ~/.cubrid.sh ulimit -c unlimited
We perform shell_long
test twice a week (controlled through like crontab) and perform code coverage test for monthly. Crontab task for shell_long test is as below:
#######################################################################
job_shell_long.service=ON
job_shell_long.crontab=* 0 11 ? * SUN,THU
job_shell_long.listenfile=CUBRID-{1}-linux.x86_64.sh
job_shell_long.acceptversions=10.0.*.0~8999,10.1.*,10.2.*
job_shell_long.package_bits=64
job_shell_long.package_type=general
job_shell_long.test.1.scenario=shell_long
job_shell_long.test.1.queue=QUEUE_CUBRID_QA_SHELL_LONG_LINUX
#######################################################################
When the build server has a new build and meet the conditions of the crontab task, shell_long
regression test will be started. If there is something wrong and need to run shell_long test again, you need send a test message by manual.
$ cd ~
$ sh start_test.sh &
$ tail -f nohup.out
Login message@192.168.1.91
and send test message.
By default, it will find the ~/CTP/conf/shell_template_for_shell_long.conf
to execute. Otherwise it will find the ~/CTP/conf/shell_template.conf
sender.sh QUEUE_CUBRID_QA_SHELL_LONG_LINUX http://192.168.1.91:8080/REPO_ROOT/store_01/10.2.0.8268-6e0edf3/drop/CUBRID-10.2.0.8268-6e0edf3-Linux.x86_64.sh shell_long default
-
Open QA homepage, then navigate as below to find shell_long test result.
If some test shows 'NO RESULT', we need investigate reasons and resolve it. -
In above picture, the figures with red color mean number of failures. Click it to open verification page. Then follow the same way as SHELL Test's to verify all failures. Both
Test Rate
andVerified Rate
should be 100%.
Code coverage test starts on the last Sunday of each month. You can find the setting from http://qahome.cubrid.org/qaresult/job/job.conf.
#######################################################################
job_codecoverage.service=ON
job_codecoverage.crontab=0 1 10 ? * 7L
job_codecoverage.listenfile=cubrid-{1}.tar.gz
job_codecoverage.acceptversions=10.2.*
job_codecoverage.package_bits=64
job_codecoverage.package_type=general
job_codecoverage.test.1.scenario=gcov_package
job_codecoverage.test.1.queue=QUEUE_CUBRID_QA_CODE_COVERAGE
#######################################################################
job_coverage_test.service=ON
job_coverage_test.crontab=0/7 * * * * ?
job_coverage_test.listenfile=CUBRID-{1}-gcov-linux.x86_64.tar.gz
job_coverage_test.listenfile.1=cubrid-{1}-gcov-src-linux.x86_64.tar.gz
job_coverage_test.acceptversions=10.2.*
job_coverage_test.package_bits=64
job_coverage_test.package_type=coverage
job_coverage_test.test.15.scenario=shell_long
job_coverage_test.test.15.queue=QUEUE_CUBRID_QA_SHELL_LONG_LINUX
-
Login
message@192.168.1.91
, using thesender_code_coverage_testing_message.sh
script to send a code coverate test message.$ cd ~/manual $ sh sender_code_coverage_testing_message.sh Usage: sh sender_code_coverage_testing_message queue url1 url2 category
queue:
queue name
url1:
gcov build package url
url2:
gcov source package url
catecory:
shell_long -
$ cd ~/manual $ sh sender_code_coverage_testing_message.sh QUEUE_CUBRID_QA_SHELL_LONG_LINUX http://192.168.1.91:8080/REPO_ROOT/store_01/10.2.0.8270-c897055/drop/CUBRID-10.2.0.8270-c897055-gcov-Linux.x86_64.tar.gz http://192.168.1.91:8080/REPO_ROOT/store_01/10.2.0.8270-c897055/drop/cubrid-10.2.0.8270-c897055-gcov-src-Linux.x86_64.tar.gz shell_long Queue:QUEUE_CUBRID_QA_SHELL_LONG_LINUX Build URL:http://192.168.1.91:8080/REPO_ROOT/store_01/10.2.0.8270-c897055/drop/CUBRID-10.2.0.8270-c897055-gcov-Linux.x86_64.tar.gz Source URL:http://192.168.1.91:8080/REPO_ROOT/store_01/10.2.0.8270-c897055/drop/cubrid-10.2.0.8270-c897055-gcov-src-Linux.x86_64.tar.gz Category:shell_long Message: Message Content: Test for build 10.2.0.8270-c897055 by CUBRID QA Team, China MSG_ID = 190705-165519-836-000001 MSG_PRIORITY = 4 BUILD_ABSOLUTE_PATH=/home/ci_build/REPO_ROOT/store_01/10.2.0.8270-c897055/drop BUILD_BIT=0 BUILD_CREATE_TIME=1551930752000 BUILD_GENERATE_MSG_WAY=MANUAL BUILD_ID=10.2.0.8270-c897055 BUILD_IS_FROM_GIT=1 BUILD_PACKAGE_PATTERN=CUBRID-{1}-gcov-Linux.x86_64.tar.gz BUILD_SCENARIOS=shell_long BUILD_SCENARIO_BRANCH_GIT=develop BUILD_SEND_DELAY=10382567 BUILD_SEND_TIME=1562313319834 BUILD_STORE_ID=store_01 BUILD_SVN_BRANCH=RB-10.2.0 BUILD_SVN_BRANCH_NEW=RB-10.2.0 BUILD_TYPE=coverage BUILD_URLS=http://192.168.1.91:8080/REPO_ROOT/store_01/10.2.0.8270-c897055/drop/CUBRID-10.2.0.8270-c897055-gcov-Linux.x86_64.tar.gz BUILD_URLS_1=http://192.168.1.91:8080/REPO_ROOT/store_01/10.2.0.8270-c897055/drop/cubrid-10.2.0.8270-c897055-gcov-src-Linux.x86_64.tar.gz BUILD_URLS_CNT=2 BUILD_URLS_KR=http://192.168.1.91:8080/REPO_ROOT/store_01/10.2.0.8270-c897055/drop/CUBRID-10.2.0.8270-c897055-gcov-Linux.x86_64.tar.gz BUILD_URLS_KR_1=http://192.168.1.91:8080/REPO_ROOT/store_01/10.2.0.8270-c897055/drop/cubrid-10.2.0.8270-c897055-gcov-src-Linux.x86_64.tar.gz MKEY_COVERAGE_UPLOAD_DIR=/home/codecov/cc4c/result MKEY_COVERAGE_UPLOAD_IP=192.168.1.98 MKEY_COVERAGE_UPLOAD_PWD=****** MKEY_COVERAGE_UPLOAD_USER=codecov Do you accept above message [Y/N]: Y
-
Go to QA homepage and find the 'code coverage' node in the left area, click the link of latest result.
-
There is a coverage rate of lines. Its coverage rate of lines is usually in 40%~42%.
Please refer to report issues of shell test
.
You can refer to http://jira.cubrid.org/browse/CBRD-21989.
It is necessary to add such information: Test Build
,Test OS
,Description
,Repro Steps
,Expected Result
,Actual Result
and Test Cases
.
Sometimes we need save database files and logs to analyze this issue.
Here are examples you can refer to.
http://jira.cubrid.org/browse/CBRD-22097
http://jira.cubrid.org/browse/CBRD-21772
We can report crash issue though tools:
There are a lot of backup files when we report crash issues,once these issus have been closed,we need to delete them.
$ cd ~/do_not_delete_core/
$ ls
10.2.0.7925-616b134_20180521-143021.tar.gz AUTO_10.2.0.7773-e468e09_20171225_063924.tar.gz AUTO_10.2.0.8038-b6e1d4b_20181010_152229.tar.gz readme.txt
When we execute shell_long test,server appears crash or fatal error,the current db and other important information will be save in ~/ERROR_BACKUP,you need check it for each build and clear them in time.
$ cd ~/ERROR_BACKUP/
$ ls
AUTO_10.2.0.8038-b6e1d4b_20181010_152229.tar.gz AUTO_10.2.0.8254-c015eb2_20190214_005631.tar.gz AUTO_SERVER-START_10.2.0.8254-c015eb2_20190213_023728.tar.gz
AUTO_10.2.0.8060-689ccdd_20181019_014026.tar.gz AUTO_10.2.0.8329-51e235e_20190413_151229.tar.gz AUTO_SERVER-START_10.2.0.8362-fbf9d84_20190526_055005.tar.gz
AUTO_10.2.0.8060-689ccdd_20181019_040434.tar.gz AUTO_10.2.0.8349-bb21e2d_20190428_190725.tar.gz AUTO_SERVER-START_10.2.0.8362-fbf9d84_20190526_055233.tar.gz
AUTO_10.2.0.8107-a05cfaa_20181028_053656.tar.gz AUTO_10.2.0.8362-fbf9d84_20190526_060734.tar.gz
AUTO_10.2.0.8254-c015eb2_20190213_025432.tar.gz AUTO_SERVER-START_10.2.0.8254-c015eb2_20190213_020024.tar.gz
The directory structure, naming rule, and convention rules of shell_long test case are definitely same as that of shell test cases. Please refer to 5. Shell Case Standards to write test cases.