So this is the only thing I see on failed build. When running npm scripts on a cli, you usually see more than the exit status. Is there some option to view the entire cli output instead of this pseudo log?
I contacted support and was told to cat the debug log in order to see the output.
#!/bin/bash
set -ex
cat $(find $HOME/.npm/_logs -name '*-debug.log')
Related
I have a CI stage with the following command, which has to be executed remotely and checks if the mentioned file exists, if yes it creates a backup for it.
script: |
ssh ${USER}#${HOST} '([ -f "${PATH}/test_1.txt" ] && cp -v "${PATH}/test_1.txt" ${PATH}/test_1_$CI_COMMIT_TIMESTAMP.txt)'
The issue is, this job always fails whether the file exists or not with the following output:
ssh user#hostname '([ -f /etc/file/path/test_1.txt ] && cp -v /etc/file/path/test_1.txt /etc/file/path/test_1_$CI_COMMIT_TIMESTAMP.txt)'
Cleaning up project directory and file based variables
ERROR: Job failed: exit status 1
Running the same command manually, just works fine. So,
How can I make sure that this job succeeds as long as command logic is executed successfully and only fail incase there are some genuine failures?
There is no way for the job to know if the command you ran remotely worked or not. It can only know if the ssh instruction worked or not. You can force it to always succeed by appending || true to any instruction.
However, if you want to see and save the output of your remote instruction, you can do something like this:
ssh user#host command 2>&1 | tee ssh-session.log
My Cloud build fails with a timeout on npm test, and no useful information is sent to stdout. A complete log can be found in a file, but I couldn't find a way to ssh in the cloud build environment.
Already have image: node
> project-client#v3.0.14 test
> jest
ts-jest[versions] (WARN) Version 4.1.0-beta of typescript installed has not been tested with ts-jest. If you're experiencing issues, consider using a supported version (>=3.8.0 <5.0.0-0). Please do not report issues in ts-jest if you are using unsupported versions.
npm ERR! path /workspace/v3/client
npm ERR! command failed
npm ERR! signal SIGTERM
npm ERR! command sh -c jest
npm ERR! A complete log of this run can be found in:
npm ERR! /builder/home/.npm/_logs/2020-11-09T07_56_23_573Z-debug.log
Since I have no problem running the tests locally, I'd like to see the content of that 2020-11-09T07_56_23_573Z-debug.log file to hopefully get a hint at what might be wrong.
Is there a way to retrieve the file content ?
Ssh in a cloud build environment?
Get npm to print the complete log to stdout?
Some way to save the log file artifact to cloud storage ?
I had a similar issue with error management on Gitlab CI and my workaround is inspired from there.
The trick is to embed your command in something that exit with a return code 0. Here an example
- name: node
entrypoint: "bash"
args:
- "-c"
- |
RETURN_CODE=$$(jtest > log.stdout 2>log.stderr;echo $${?})
cat log.stdout
cat log.stderr
if [ $${RETURN_CODE} -gt 0 ];
then
#Do what you want in case of error, like a cat of the files in the _logs dir
# Break the build: exit 1
else
#Do what you want in case of success. Nothing to continue to the next step
fi
Some explanations:
echo $${?}: the double $ is to indicate to Cloud Build to not use substitution variables but to ignore it and let Linux command being interpreted.
The $? allows you to get the exit code of the previous command
Then you test the exit code, if > 0, you can perform actions. At the end, I recommend to break the build to not continue with erroneous sources.
You can parse the log.stderr file to get useful info in it (the log file for example)
I am trying to fail a build in gitlab CI and get email notification about it.
My build script is this:
echo "Listing files!"
ls -la
echo "##########################Preparing build##########################"
mkdir build
cd build
echo "Generating make files"
cmake -G "Unix Makefiles" -D CMAKE_BUILD_TYPE=Release -D CMAKE_VERBOSE_MAKEFILE=on ..
echo "##########################Building##########################"
make
I have commited the code that breaks build. However, instead of finishing, build seems to be stuck in "running" state after exiting make. Last line is:
make: *** [all] Error 2
I also get no notifications.
How can i diagnose what is happening?
Upd.: in runner, following is repeated in log:
Submitting build <..> to coordinator...response error: 500
In production.log and sideq.log of gitlab_ci, following is written:
ERROR: Error connecting to Redis on localhost:6379 (ECONNREFUSED)
Full message with stacktrace is here: pastebin.
I have the same problem, i can help you with a workaround but im trying to fully fix it.
1- most of the times he hangs but the jobs keeps on going and actually finishes it, you can see the processes inside the machine, example: in my case it compiles and in the end it uses docker to publish the build, so the process docker doesn't exist until he reaches that phase.
2- to workaround this issue you have to make the data persistent and "retry" the download over and over again until he downloads everything he needs.
PS: stating what kind of OS you are using always helps.
I was wondering if I could use bamboo's SSH task to run a script (this kicks off a small java message injector).
Then grep the logs for ERRORS. If any ERROR is present I would like to fail the build.
Something like this:
Is this a Bash question or is it really about Bamboo? Here is the Bash problem answer:
If you run
[[ ! $(grep ERROR /a/directory/log/*) ]]
the script will exit with an error if it finds the word "ERROR" anywhere in the files.
Bamboo should detect the task execution as failed.
(Note that if Bash is not the default shell on your target system you may need a #!/bin/bash on top of the script file.)
In jenkins post build action I configured Execute shell script on remote host using ssh
ssh site 10.32.25.66, command:
cd $HOME/appsadm/bin; ./ims-carte-stop
and i again modified
cd /HOME/appsadm/bin; ./ims-carte-stop.*
I tried both these commands and Build is successful, but I see in console output in Jenkins after, that it is not executing my script. I am getting ssh exit status 1 error.
In my winscp my script (ims-carte-stop) in this location home/appsadm/bin.
Please tell me if I am doing aything wrong.
My intention is to stop my server from jenkins automatically whenever the build success.
This may be a typo in your question, but:
You said your ims-carte-stop script is in:
/home/appsadm/bin
whereas your script is doing:
cd $HOME/appsadm/bin
or
cd /HOME/appsadm/bin
Looking at the paths, I am going to assume you are using a UNIX-flavoured OS (Linux, BSD, OSX).
UNIX paths are case sensitive. Your script should be calling:
cd /home/appsadm/bin
Note that the word "home" is all small letter not capitals. Also, using $ makes it a variable, which I don't think you want.