As of yesterday, 3/12/2014 my cron job for my free python 3.3, mongodb, cron app on Openshift stopped running. Everything was working fine until late afternoon. I can still run the cron job manually without error.
The execution bit is set. I wouldn't think I have run out of processing time since I can run it manually. I'm not sure if it's related to this issue:
Openshift app push error 98
Both issue started at the same time. Any ideas?
Related
I have my GCP machine that is scheduled to switch on and run from 6hrs in the morning to 12hrs. I am trying to automate the script once the machine is on, so I am scheduling a job on 6:30 hrs.
But what I observed is that UIPath Robot doesn't automatically get connected unless I open an RDP session and then manually open UI Path Assistant to connect it with orchestrator.
I also tried to create a bat script and schedule it with windows startup. But that fails to run as well. To verify my startup script I added two commands:
<verification script sending mail on startup> && <UIPath Script>
I tested this script manually in RDP session and they did run sequentially as expected. But when I added it to auto-scheduling in gcp, only the first script actually ran (<verification script sending mail on startup>) and the second one (<UIPath Script>) just didn't.
Moreover I did put the bat file in Windows Startup Folder but again that also required me to sign into the system using RDP session for it to run.
From my understanding, the UIPath Script (XAML) in order to run needs the system to be switched on 24/7 and then it connects to orchestrator and then we can schedule the process based on the time, but then my maching will be idle for lots of ineffective hours.
Secondly, going by the .bat file to be as a startup script also fails as in order to run the script, it needs the user to be logged in which doesn't happen. Though other scripts such as .py files run smoothly so I lose here as well.
Finally scheduling my bat file through windows startup feature does work but then it's also of no use as for it to start also I need to open an RDP session and log in.
Does anyone know the workaround for this?
I've built my container multiple times successfully, as noted by the image below. Each time it remains on 99% for > 20+ mins AFTER saying 'Finished: SUCCESS' in the logs. It never makes it past this. I cant kick off the deploy phase until the build registers completion. Is there a way to get past this hang?
I've got no notable errors in the console. The build is based on the registry.ng.bluemix.net/ibmnode:latest image, runs an apache2 server with some Node.js processes that run during the build phase. And lastly, it kicks off a bash script to run apache2 in the foreground.
I just checked my toolchain and wasn't able to reproduce this problem. Please try again, it might have been a transient issue with the toolchains.
If the problem persists, it might have to do with how you have your build script setup. If you are spawning processes and leaving them running, that could be stopping the build from finishing.
I want to run an OpenShift application that takes a few hours to run. Once the application finishes (exit 0) I don't want it run again via a restart. I just want to look at the logs to see what it did.
One way of achieving this is for the application to stop itself once it's done processing using the OpenShift API. I'm hoping there is an easier way than this.
The application uses node.js
The use case is I'm running back tests for a trading application. There is no UI to the application. It's just processing historical trades stored in a large file.
I am calling a perl script using my hudson job.
This script is responsible for getting data from cvs and svn and then running
an operation on this data.
Now during the weekend this script just didn't stop running.
This happened because a particular cvs location was locked and the script was waiting for the locks to be released.
How can I make hudson notify me in such scenarios?
Check out the build timeout plugin.
How can I check if a build is running for more than X minutes? Does Hudson have some XML API for it? I want the API to be independent of the job, so I can call a URL from my remote machine to check if Hudson is executing any build or any job for more than x minutes.
It may be the case that the job that Hudson is executing is stuck or hanged. How can I identify that?
My idea is if a build is taking too long I will just restart the system or kill all its processes
I will call a URL, say http://myhudsonserver/something, and I should get some XML by which I can make out.
You could use the build timeout plugin.
You may find the Hudson Remote Access API of use.