My jobs build at the same time every day.
But I don’t know why some job took a long time.
Does anyone know why this happens?
enter image description here
After that I thought of looking at the full log output from the console.
Confirm when something went wrong
But when clicking full log, not all logs are listed.
enter image description here
Jenkins stored log does not show time.
Or is there any way to read timestamps file?
Related
I have a stored procedure created some time ago, actually in 2016 and it is set to run every year in the first day. The procedure automatically ran in 2017.01.01 but now in 2018 this did not happened automatically and i had to trigger it manually. Do you have any hints of what can i check to see what went wrong?
Below you can find the requested details:
DB2 server operating system: Linux ;
DB2 version: 10.1.3.0 ;
Scheduler: I am using cron.
I have checked the ADMIN_TASK_LIST and the END_TIME column is NULL, so from my point of view this should never stop running;
2.On first of January there is no entry for this in 2018, only in 2017;
I will check the files and let you know ;
I don't find any ATS_TASK_LIST, do you mean ADMIN_TASK_LIST?
I have checked db2diag and there was a space error identified but as per my colleagues feedback should have nothing to do with this. Also, i want to mention that yesterday we have created a test task which was set to be executed today at a certain hour and was not triggered. I have attached an image with the TEST task details in order for you to see if there is anything set wrong. Also, here is the error that i was talking about:"The Event Monitor "DB2DETAILDEADLOCK" has reached its file capacity"
TEST TASK IMAGE
It appears you don't know what schedules your job.
If cron is the scheduler (as you originally wrote), then most likely your question has nothing to do with Db2, see checks at the end of this answer. If you lack the skills with cron, ask your linux admin or study the docs and there's plenty of online info.
If the Db2 Administrative Task Scheduler (ATS)is the scheduler (as you subsequently imply from your comments), then the problem can be with Db2. Cron has nothing to do with the ATS, they are unrelated. So then your question is "why did ATS not run my job". This is also often asked.
Suggested checks for ATS jobs.
Check if the job is still present in ADMIN_TASK_LIST with correct definition , especially the END_TIME column.
Check if there is an entry in ADMIN_TASK_STATUS for this job on 01-January and study the entry for details of why the job failed
If the job is present in ADMIN_TASK_LIST with correct definition, and if there is no entry for the job in ADMIN_TASK_STATUS then check if the db2acd is still running, and check if db2acd has run out of memory (evidence in db2diag). I have seen that stopping db2acd cures this problem (Db2 watchdog will immediately restart db2acd).
If you are still stuck then edit your question to show the ATS_TASK_LIST entry and the ADMIN_TASK_STATUS entry for the failed run. Take care to give accurate responses to requests for clarification.
============================================
If cron is the scheduler, the advice below applies:
It is likely that your question is "why did cron not run my job".
That is a frequently asked question, so you can research that (there are many pages giving advice on that topic).
See https://unix.stackexchange.com/questions/207/where-are-cron-errors-logged
Check the crontab entry to ensure it is still valid.
Check the cron entry command (usually a script) still exists, and has
the correct permissions and that it can run with the
account specified .
Investigate the cron log file (if that is still available) and study that file
for clues . This will show if cron attempted to start the job, and what the exit code was.
When cron fails to run a job and that job had some output, cron emails the crontab owner with the output of the failed job.
But this can depend on the exact command that is in the cron entry
(in case it redirects stdout and stderr to some file instead, in
which case you should consult that file).
If the cron entry does not redirect stdout/stderr, and the script/command gives output when run then refer to the mailbox of the cron owner to see what emails arrived on first of january.
I run Jenkins as a standalone build machine.
It works fine, I have already created multiple pipelines.
But it displays two different time for the same builds in Stage View and Build history.
The difference between the time is exactly 2 hours.
Look at the attached screenshot for more demonstrative illustration of the problem.
How I can fix this?
Looks like the build time under build history is getting picked from Jenkins server time and the stage view time is your system's local time. Looks like a bug to me.
Temporary fix can be to match your timezone.
On Jenkins script console https://jenkins-servername/script
Run
System.setProperty('org.apache.commons.jelly.tags.fmt.timeZone', 'America/New_York')
Set it to match your timezone. Read
Had the same problem today. I solved it by changing the 'User Defined Time Zone'.
Current User > Configuration > User Defined Time Zone > Select correct time zone
The problem comes when the time zone configured in the user profile differs from de one configured in the cron of the pipeline.
This is a broad question, so any answers are deeply appreciated. I need to continually log the size of several build files (in this case some CSS and JS files), preserve this log and ideally show it as a dashboard in Jenkins.
I know that I can setup a cron job and execute a bash script to grab the files and log their size, but I'm not sure where this file would live and how to display it. Ideally the result would be a dashboard plot or bar graph over time.
Thanks.
P.S. I'm open to other logging suggestions, but Jenkins seems like the appropriate system to do this in.
Update: this isn't perfect but it works. Google Spreadsheets has a simple API for posting data, so this can work as an endpoint for any script you want to write that logs your data.
It's not a Jenkins solution, but gets the job done.
In my search leading up to this, I did come across JMeter, and the Performance Plugin for Jenkins, which were contenders for a possible solution.
I'm having a problem with a job in collabnet edge. I created a blank repository, then did a load from a dump file. There was an issue during the load (finally figured out that I had run out of disk space) and due to this, a job got stuck.
So, here's what's happening:
The load appeared to finish, the job isn't shown on the list.
The data was not loaded correctly (no space...but took a while to figure that out)
So I deleted the repository, added disk space and then tried to reload, but I get a message saying I can't because there is already a job running.
A dump file is already set to be loaded. Only one load may be scheduled at a time; progress can be monitored on the Jobs screen.
As mentioned, there is no job listed as being in progress. The repo it was loading has been deleted.
How do I clear out this stuck job?
Check, if there is a file in the following directory:
csvn/data/dumps/yourRepositoryName/load
Delete the file and then reload your dump-file.
Check if there is a job progress for this repository logged under csvn/data/logs/temp
If present remove it and then re-schedule
I am having trouble getting email-ext working. I've looked at the system log ("Manage Jenkins"/"System Log"), and it shows the job succeeded in running, but nothing about how email works.
Is there some way to turn up logging so I can watch the emailing portion of either default or email-ext firing? Maybe I can tell why email is not working if I can see what it's trying to do.
Edit: We're running Jenkins 1.484.
you can do this:
go to configure system-system log,and add a new recorder
Choose a name that makes sense to you.
configure loggers and their levels whose output you'd like to collect
you can look there:https://wiki.jenkins-ci.org/display/JENKINS/Logging