Add Partial Output to Tango, and functionality to localDocker VMMS

[victorhuangwq]

Description

Adds partial output functionality by:

• Adding Endpoint on restful_tango that takes in JobID
• Adds getPartialOutput function to tango.py, which also does error handling and calls on the VMMS to get partial output
• Adds functionality to getPartialOutput for localDocker, mainly as this is likely the most use VMMS for Tango right now.
• Adds calling of endpoint via tango-cli for easy testing
• Adds job8-10 which is a grading test case that takes at least 30 seconds to run, which makes the rest of the testing easier by hand. 8 is a C version, 9 is Python and 10 is bash.

How to test

• Have a fully functioning Tango, following instructions here, using local docker as VMMS https://docs.autolabproject.com/installation/tango/#development-installation
• For the rest of the testing, I'm assuming that you left the Tango Key as default of just test (DEFAULT_KEY = "test" in config.py`)

Test situation when jobId doesn't exits / is no longer live or on the queue (could be completed)

• Run python clients/tango-cli.py --getPartialOutput -k test --jobid 1
• Output should be {"statusId": -1, "statusMsg": "getPartialOutput request failed: Invalid job id"}

Test situation when a VMMS doesn't have getPartialOutput Implemented

• Change localDocker's getPartialOutput function to getPartialOutput2
• Run python clients/tango-cli.py -P 3000 -k test -l assessment10 --runJob clients/job10/ --image autograding_image(this submits job10 for grading)
• Output should be something like {"statusId": 0, "statusMsg": "Job added", "jobId": "1"}
• Run python clients/tango-cli.py --getPartialOutput -k test --jobid 1
• Output should now be {"statusId": -1, "statusMsg": "getPartialOutput request failed: 'LocalDocker' object has no attribute 'getPartialOutput'"}
  ( make sure to change the function getPartialOutput name back)

Test getting partial output

• Run python clients/tango-cli.py -P 3000 -k test -l assessment10 --runJob clients/job10/ --image autograding_image(this submits job10 for grading) to submit another job
• Run python clients/tango-cli.py --getPartialOutput -k test --jobid 1 (depending on the job id)
• Output should be {"statusId": 0, "statusMsg": "Partial output obtained", "output": "bash hello.sh\r\nHello to sleep job\r\nHello 1 times\r\nHello 2 times\r\nHello 3 times\r\nHello 4 times\r\nHello 5 times\r\nHello 6 times\r\nHello 7 times\r\nHello 8 times\r\nHello 9 times\r\nHello 10 times\r\nHello 11 times\r\n"}

Test situation when the Job is still on the queue

• Run python clients/tango-cli.py -P 3000 -k test -l assessment10 --runJob clients/job10/ --image autograding_image(this submits job10 for grading) to submit another job 3 times (if you are on default, as by default settings, Tango only spins up at most 2 in the autograding_pool).
• Run python clients/tango-cli.py --getPartialOutput -k test --jobid 3 (depending on the job id of the last job to be submitted
• Output should initially be {"statusId": -1, "statusMsg": "getPartialOutput request failed: Job 4 is not running yet"}
• Once the first two jobs are done, running python clients/tango-cli.py --getPartialOutput -k test --jobid 3 should now get you the partial output

[michellexliu]

lgtm

[damianhxy]

Followed testing plan, code looks good

lgtm