-
Suggestion
-
Resolution: Unresolved
-
None
-
None
-
None
Issue Summary
Updating an AWS instance(Cloudformation) to a new version or deploying the instance to a new AMI would ignore the custom script function in the DC environment. The DC System Administrator had to manually copy over the script(and build it) in the new deployment. The custom script expected to monitor the nodes and perform cronjobs in nodes(i.e: removing logs - the node size limited to 8GB)
It is expected that the custom scripts have a default directory in Bitbucket_Home(i.e: /media/atl/bitbucket/shared/node_scripts). The nodes are able to access and copy the custom script from the mentioned directory and automate/run the script in each node's environments.
- mentioned in
-
Page Failed to load
[SCALE-22] Custom script to be copied from one deployment to another deployment
Remote Link | New: This issue links to "Page (Confluence)" [ 503298 ] |
Summary | Original: Custom script to be copied from one deployment to another deployment. | New: Custom script to be copied from one deployment to another deployment |
Background:
SCALE-23enhancementSCALE-22 was born out of the need for
SCALE-23mitigations.This particular change this would allow a BB admin to reliably execute some small customizations after a node initializes; for example: set up a cron job that purges log folders or install Monit so we get notified if something breaks.... but these are not permanent fixes if the node terminates.
The idea is/was as follows:
1. define a folder on the shared location, ea: /shared/custom_node_scripts2. admins can place scripts or packages in here;
2. When a new node emerges after doing it's Ansible script, etc...it copies the scripts from the shared location and runs them from a local folder
Extra Requirment: These scripts run once, and The order of execution is determined by renaming them with a number.
I feel this fairly easy enhancement would benefit a lot of customers.