+def verify_rasd(server, assoc_cn, cn, virt, list, rasd): + try: + data = assoc.AssociatorNames(server, + assoc_cn, + cn, + virt, + InstanceID=list) + except Exception: + logger.error(Globals.CIM_ERROR_ASSOCIATORNAMES % cn) + return FAIL
It's possible for the AssociatorNames call to return no instances without throwing an error. You'll want to verify that data has at least one element.
+ + if data[0]['InstanceID'] not in rasd:
By using data[0], you're relaying on the CIMOM to return the instance corresponding to test_dom first in the list. This might not always be the case. Instead, you should step through the items in data and select the instance or instances that correspond to test_dom. If you don't find an instance, you should return an error. Also, you're checking the value of InstanceID against all the possible rasd InstanceIDs for test_dom. If the provider returns the MemRASD that corresponds to test_dom when it should have returned the ProcRASD, this check will return a false positive. To fix this, you could change rasd from an array to a list. Something like the following. You already have the pool classname in the verify_rasd() call, so you can use it as a key value to get the corresponding RASD InstanceID. pool_to_ rasd = { {'MemoryPool' : "%s/mem" % test_dom}, {'ProcessorPool' : "%s/proc" % test_dom}, {'DiskPool' : "%s/%s" %(test_dom, test_disk)}, {'NetworkPool' : "%s/%s" % (test_dom, test_mac)} }
+ logger.info ('ID %s' % data[0]['InstanceID']) + logger.error("InstanceID Mismatch") + logger.error("Returned %s error" % data[0]['InstanceID'])
Then here, you can print the ID value AssociatorNames returned versus the expected value.
+ + if options.virt == 'LXC': + pool = { "MemoryPool" : {'InstanceID' : "MemoryPool/0"} } + rasd = [ "%s/mem" % test_dom ] + else: + pool = { "MemoryPool" : {'InstanceID' : "MemoryPool/0"}, + "ProcessorPool" : {'InstanceID' : "ProcessorPool/0"}, + "DiskPool" : {'InstanceID' : diskid}, + "NetworkPool" : {'InstanceID' : "NetworkPool/%s" % test_network }}
Can you fix this so that the line is 80 characters in length.
+ rasd = [ "%s/mem" % test_dom, + "%s/proc" % test_dom, + "%s/%s" %(test_dom, test_disk), + "%s/%s" % (test_dom, test_mac) ]
+ for k, v in pool.iteritems(): + status, inst = get_instance(options.ip, k, v, options.virt)
If get_instance() returns an error, you should break out of the loop here. The test is only returning the status of the last iteration through the loop. If a previous iteration fails, but the last iteration succeeds, the test returns a false positive.
+ + status = verify_rasd(options.ip, "ResourceAllocationFromPool", + k, options.virt, inst.InstanceID, + rasd)
Same here - you'll need to trap the status and then break from the loop in case of error. -- Kaitlin Rupert IBM Linux Technology Center kaitlin@linux.vnet.ibm.com