/
Phase-2:30. New Requests

Phase-2:30. New Requests

#

Task

Details

Questions

Status

#

Task

Details

Questions

Status

1

Deletion of temporary files from Jira

During the process of transferring files from Imaging server to NETX we store the files temporarily in JIRA’s local “temp“ directory. When the transfer is done we have to remove these temporary files to prevent Jira server from running out of space.

 

open

2

Make app more cluster friendly

If possible files should be written to a shared storage which is accessible across nodes and not local storage on each node

Nodes can go down or can be brought down due of maintenance. Now if the node goes down while it is processing the requests then there should be a way for the node to either resume the work from where it left off when the node went down or an existing node should take up the responsibility to either restart the processing of failed request or resume from where the previous node(node which went down) left it off.

Info: Artic uses sticky session with their load balancer(a new session is established if the node goes down or if the browser is closed by the user)

 

open

3

Can we update the app such that the node that receives the file is not required to be the one to process it?

Can we update the app such that the node that receives the file is not required to be the one to process it? this will be more in line with our updated cluster architecture. we have two application nodes - node01 and node02 and we are seeing the majority of imaging activity done on node01

 

 

4

Create one sub-task per object id (multi-object type 1) for other projects other than imaging requests project in Artic Jira

Check if we can add functionality to the app such that regardless of the Jira Project, if we launch requests from CITI and Object ID contains multiple values (multiObjectType1 single form), can we create sub-tasks per Object ID and then pull data from CITI. We currently create sub-tasks in the Imaging project based on View, Light Type, and the custom field that Gaurav added in phase 2. However, as we scale this application, we have other teams that do not need the View, Light Type, etc. sub-tasks but would like to offer the ability to use the single form, multi object functionality

 

open

4

Push code to Github repo

Push the latest code to Github repo provided by Atticus

 

open

5

Theoretically, if we were to add the Object ID field to the Object (not in CITI) request form, then populate it with a comma separated list, would it still create sub-tasks in the same way as Object (in CITI) creates sub-tasks based on objects x views, light type etc.or does it need additional information from CITI in order to trigger the sub-task generation?such as CITI request type ID

 

 

 

 

Note from Rafe:-

We periodically freshen our test instance by copying our live jira instance over to test and in so doing would wind up copying uploaded images and perhaps other metadata in the process.  Assuming the imaging app is enhanced to the extent that it will be more cluster friendly, that it will only store images and related data in a way that all nodes can access and establish a mechanism by which the app can keep track of what's been processed and what is pending so that if a node goes down another can take over, I have the following concern: Copying live to test might include pending imaging app tasks which we wouldn't want the test instance to actually process. To address this, it would be nice if there could be a way to suppress processing
on the test instance.  For example, we apply the following environment setting in setenv.sh on the test instance to prevent sending or retrieving email:DISABLE_NOTIFICATIONS=" -Datlassian.mail.senddisabled=true -Datlassian.mail.fetchdisabled=true -Datlassian.mail.popdisabled=true"Is something like this to suppress processing within the imaging app possible?Also, assuming the requested cluster friendly enhancements are made, can there be a simple and quick way to clear out (remove) any pending work that get's copied over from live to test?  This would allow us to remove any pending work and then re-enable the app for testing without worrying about the test instance processing production data that's already being handled on the live instance.