southampton/cortex

A VM orchestration system for VMware, Infoblox, ServiceNow, Puppet, TSM, Rubrik, Red Hat Satellite 5 / 6. Under heavy development, and somewhat still heavily tailored to University of Southampton.

cortex's Issues

Add notice to Puppet section saying that it's only for new Puppet builds (e.g. punch.puppet/judy.puppet rather than rhel.puppet)

NeoCortexLib should have method for updating AD Computer object attributes (for populating bginfo, for example).

Sandbox VM: Add Environment AND Network

Let the user choose what network and environment for the VM.

Add ServiceNow environment and description global variables into Puppet

Switch graphing to another library

http://c3js.org/samples/chart_donut.html

Port trackit's global error handler into cortex

Create CSV download on other table pages like vmware data

pyro4 cant send custom objects across rpc

By default, custom classes are serialized into a dict. They are not 
deserialized back into instances of your custom class. This avoids 
possible security issues. An exception to this however are certain
classes in the Pyro4 package itself (such as the URI and Proxy 
classes). They are deserialized back into objects of that certain
class, because they are critical for Pyro to function correctly.

See: https://pythonhosted.org/Pyro4/clientcode.html#changing-the-way-your-custom-classes-are-de-serialized

Seems there is a way around this, but we might not want to do it.

[TODO] Document manual procedure, for when Cortex is unavailable

Update Git!

Workflow: Sandbox

Add authentication to ENC API

Pull in Environment and "Is Virtual" from ServiceNow

Add puppet enc groups/netgroups support

On VM create update the vmware_cache for the new VM

This is ESSENTIAL for the Puppet UUID auto join system.

Add in task finished page with VM details

Wait for VM power up before doing cache update in workflows

Make cortex-enc-wrapper use a config file

Decide on file format? Use Py-file (like elsewhere) or INI (as it's easier)

Add icon legends on claytons list pages

expire_old_tasks should cancel events within tasks that it expries

Pull in other CIs from ServiceNow (e.g. SAN switches, etc.)

Remove all CDN requirements so Cortex functions without internet access

Write the standard server workflow

Ping ip addresses in case infoblox doesnt know if the IP address is actually in use

@claytonpeters Gary raised this issue, discuss with you later.

Managing/Updating VMware 'Notes' field as YAML

Sandbox VMs: Create service now entries for play vms as well

CSV export of allocated names

Possibly an API for a scheduled task, as well as a download link

Import existing spreadsheet into system names

Ability to add "other" (or legacy) names into the systems

Create production VM for Cortex

timeout jobs and mark them as timed out

Integrate vmstat

Caching database table

Need a function in neocortex helper to put CMDB sys_id in to database

Add Puppet ENC Support

Add feature tags to classes

Print out row or integrate VM charging spreadsheet

Add socket and core sliders to workflows

And when the defaults are chosen it sets 1-socket multiple-cores

Finish Puppet in-app help

Create status page for all tasks

Admin only? Per-user version?

Create a description attribute for tasks for better admin understanding of task history

Add logging to cortex-enc-wrapper

For reasons regarding why an error is happening (e.g. wrong permissions on cache files, etc.)

Check with IPtrack API regarding whether an IP has been seen in use recently

Fail in that condition? But if a VM is deleted and an IP address freed up early in a subnet because of it, won't that break?

Double-check model in ServiceNow for "Generic Linux server" (e.g.), ticket number needs going in to CI

Puppet Dashboard: Drill down in to which nodes are changed, unknown etc.

Which nodes are currently not reporting? Who knows? We should probably be able to find out!

Show allocated name(s) after allocation

when allocating names cortex redirects you to the list without actually displaying what it allocated to you.

For name(s) the allocation process should print what was just allocated...

API for service now to call

Workflow: ECS Student VMs

Consider settings per workflow

Add permissions system

Add legacy/other tabs to systems view

At the moment you can't view systems other than in the standard naming scheme.