Python Faceplant

Kustom42 · 07-01-2014, 05:27 PM

Hey Everyone,

Thanks for taking the time to look at what I'm doing here. My disclaimer: I very, very rarely code. I've done a few python and perl scripts in my day but it has never been my strength. I have completed the "learn python the hard way" tutorial so the fundamentals are there just can't seem to put it together.

This script uses vCenter API calls to get a list of all the VMs on our vCenter server. I then want to execute a shell subprocess to query against a proprietary database and compare the two to find any that exist in the vCenter API call but not in the database.

I am able to successfully execute the subprocess against the database and get the results from the API call but am having trouble getting the two to compare properly, I'm making backwards progress now from attempting multiple different methods found on google so apologize for any sloppiness.

Code:

#!/usr/bin/python

from pysphere import VIServer
import re

import shlex, subprocess, sys


def compare_vms(x, y):
        return frozenset(x).intersection(y)


vCenterServer = "REMOVED"
username = 'REMOVED'
password = 'REMOVED'

### Perform the vCenter API call, store results in vmList. Working as intended

server = VIServer()
server.connect(vCenterServer,username,password)
vmList = server.get_registered_vms()
server.disconnect()

regularVMs = []
templates = []

### Setup our query and the use shlex to split up the arguments. Working correctly.
vmsConfigured = 'mminfo -ak -q "savetime>3 days ago,pool=REMOVED" -r "vmname"'
args = shlex.split(vmsConfigured)

print args

# Execute the subprocess to query our DB. This works as intended but always displays the results to stdout. I could overlook that but would prefer to not see the entire 1000+ list returned to the CLI
query = subprocess.Popen(args)

#print query


### Sort our vmList set to pull out any VMs that are templates, I dont care about templates. This works as intended. I can do a for loop to print the regularVMs list with no errors.
for vm in vmList:
        m = re.search('\.vmt?x$', vm)
        if m.group(0) == '.vmtx':
                templates.append(vm)
        else:
                regularVMs.append(vm)


## And here is where I just fall on my face and have no idea what the best way to skin this cat is.
print "----New VMs-----" 
for vm in sorted(set(regularVMs) - set(query)):
       print vm

Any help or suggestions are very much appreciated!

This is difficult to test as it uses API calls and proprietary DB but sample data is below:

I probably need to split this or something to compare properly(havent really tried that yet)

Code:

vCenter API Call results:
----Regular VMs-----
[DATASTORE] VMNAME/VMCONFIGFILE.vmx
[DATASTORE] VMNAME/VMCONFIGFILE.vmx
[DATASTORE] VMNAME/VMCONFIGFILE.vmx

Code:

mminfo/db results:
VMNAME
VMNAME
VMNAME

dugan · 07-01-2014, 05:42 PM

So you want:

Code:

[DATASTORE] VMNAME/VMCONFIGFILE.vmx

changed to:

Code:

VMNAME

for the comparison. Is that correct?

And in the code, regularVMs is the "[DATASTORE] VMNAME/VMCONFIGFILE.vmx" list and query is the "VMNAME" list, right?

In that case,

Code:

for vm in sorted(set(regularVMs) - set(query)):
       print vm

Becomes

Code:

regularVmS = [vm.split()[1].split('/')[0] for vm in regularVMS]
for vm in sorted(set(regularVMs) - set(query)):
       print vm