Conflict Resolution: Python

For reasons explained in the Introduction to conflict resolution, we strongly recommend adopting a conflict resolution strategy that requires applications to resolve siblings according to use-case-specific criteria. Here, we’ll provide a brief guide to conflict resolution using the official Riak Python client.

How the Python Client Handles Conflict Resolution

In the official Python client, every object of the RiakObject class has a siblings property that provides access to a list of an object’s sibling values. If there are no siblings, that property will return a list with only one item. Here’s an example of an object with siblings:

bucket = client.bucket('seahawks')
obj = bucket.get('coach')
obj.siblings

# The output:
[<riak.content.RiakContent object at 0x106cc51d0>, <riak.content.RiakContent object at 0x108x1da62c1>]

So what happens if the length of obj.siblings is greater than 1, as in the case above? The easiest way to resolve siblings automatically with the Python client is to create a conflict-resolving function that takes a list of sibling values and returns a single value. Such resolution functions can be registered either at the object level or the bucket level. A more complete explanation can be found in the section directly below.

Basic Conflict Resolution Example

Let’s say that we’re building a social network application and storing lists of usernames representing each user’s “friends.” Each user will be of the class User, which we’ll create below. All of the data for our application will be stored in buckets that bear the bucket type siblings, and for this bucket type allow_mult is set to true, which means that Riak will generate siblings in certain cases—siblings that our application will need to be equipped to resolve when necessary.

The question that we need to ask ourselves at this point is the following: if a given user has conflicting lists, which list should be deemed more “correct?” What criteria should be applied? Should the lists be merged? Should we pick a list at random and deem that list correct? We’ll keep it simple here and say that the following criterion will hold: if multiple conflict lists exist, the longer list will be the one that our application deems correct. While this might not make sense in real-world applications, it’s a good jumping-off point.

Creating Our Data Class

We’ll start by creating a User class for each user’s data. Each User object will consist of a friends property that lists the usernames, as strings, of the user’s friends. We will also create a to_json method, as we’ll be storing each User object as JSON:

class User(object):
    def __init__(self, username, friends):
        self.username = username
        self.friends = friends

    def to_json(self):
        return vars(self)

Now, we can create User objects and see what they look like as JSON:

new_user = User('riakuser127', ['captheorem', 'siblingsrule572'])

new_user.to_json()
# {'username': 'riakuser127', 'friends': ['captheorem238', 'siblingsrule572']}

Implementing and Registering a Conflict Resolution Function

Let’s say that we’ve stored a bunch of User objects in Riak and that a few concurrent writes have led to siblings. How is our application going to deal with that? First, let’s say that there’s a User object stored in the bucket users (which is of the bucket type siblings, as explained above) under the key bashobunny. We can fetch the object that is stored there and see if it has siblings:

bucket = client.bucket_type('siblings').bucket('users')
obj = bucket.get('bashobunny')

print len(obj.siblings) > 1

If we get True, then there are siblings. So what do we do in that case? The Python client allows us to write a conflict resolution hook function that will be triggered any time siblings are found, i.e. any time len(obj.siblings) > 1. A hook function like this needs to take a single RiakObject object as its argument, apply some sort of logic to the list of values contained in the siblings property, and ultimately return a list with a single “correct” value. For our example case, we’ll return the value with the longest friends list:

def longest_friends_list_resolver(riak_object):
    # We'll specify a lambda function that operates on the length of
    # each sibling's "friends" list:
    lm = lambda sibling: len(sibling.data['friends'])
    # Then we'll return a list that contains only the object with the
    # maximum value for the length of the "friends" list:
    riak_object.siblings = [max(riak_object.siblings, key=lm), ]

Registering a Conflict Resolver Function

In the Python client, resolver functions can be registered at the object level, as in this example:

bucket = client.bucket_type('siblings').bucket('users')
obj = RiakObject(client, bucket, 'bashobunny')
obj.resolver = longest_friends_list_resolver

# Now, when the object is loaded from Riak, it will resolve to a single
# value instead of multiple values when both commands are executed:
obj.reload()
obj.store()

Alternatively, resolvers can be registered at the bucket level, so that the resolution is applied to all objects in the bucket:

bucket = client.bucket_type('siblings').bucket('users')
bucket.resolver = longest_friends_list_resolver

obj = RiakObject(client, bucket, 'bashobunny')
obj.reload()
obj.store()

# The resolver will also be applied if you perform operations using the
# bucket object:

bucket.get('bashobunny')
bucket.get('some_other_user')

Conflict Resolution and Writes

In the above example, we created a conflict resolver that resolves a list of discrepant User object values and returns a single value. It’s important to note, however, that this resolver will only provide the application with a single “correct” value; it will not write that value back to Riak. That requires a separate step. When this step should be undertaken depends on your application. In general, though, we recommend writing objects to Riak only when the application is ready to commit them, i.e. when all of the changes that need to be made to the object have been made and the application is ready to persist the state of the object in Riak.

Correspondingly, we recommend that updates to objects in Riak follow these steps:

Read the object from Riak
Resolving sibling conflicts if they exist, allowing the application to reason about one “correct” value for the object (this step is the subject of this tutorial)
Modify the object
Write the object to Riak once the necessary changes have been made

You can find more on writing objects to Riak, including code examples from the official Python client library, in the Developing with Riak KV: Usage section.

More Advanced Example

Resolving sibling User values on the basis of which user has the longest friends list has the benefit of being simple but it’s probably not a good resolution strategy for our social networking application because it means that unwanted data loss is inevitable. If one friend list contains A, B, and C and the other contains D and E, the list containing A, B, and C will be chosen. So what about friends D and E? Those usernames are essentially lost. In the sections below, we’ll implement an alternative strategy as an example.

Merging the Lists

To avoid losing data like this, a better strategy would be to merge the lists. We can modify our original resolver function to accomplish precisely that and will also store the resulting User object:

from riak.content import RiakContent

def longest_friends_list_resolver(riak_object):
    # We start with an empty set
    friends_list = set()

    # Then we add all the friends from all siblings to the set
    for user in riak_object.siblings:
        friends_list.update(user.data['friends'])

    # Then we make a new User object. First, we fetch the username from
    # any one of the siblings, then we pass in our new friends list.
    username = riak_object.siblings[0].data['username']
    new_user = User(username, list(friends_list))

    # Now we reuse the first sibling as a container for the merged data
    riak_object.siblings[0].data = new_user.to_json()

    # And finally we set the siblings property to include just the
    # single, resolved sibling
    riak_object.siblings = [riak_object.siblings[0]]

The drawback to this approach is the following: with a conflict resolution strategy like this, it’s more or less inevitable that a user will remove a friend from their friends list, and that that friend will end up back on the list during a conflict resolution operation. While that’s certainly not desirable, that is likely better than the alternative proposed in the first example, which entails usernames being simply dropped from friends lists. Sibling resolution strategies almost always carry potential drawbacks of this sort.

Riak Data Types

An important thing to always bear in mind when working with conflict resolution is that Riak offers a variety of Data Types that have specific conflict resolution mechanics built in. If you have data that can be modeled as a counter, set, or map, then you should seriously consider using those Data Types instead of creating your own application-side resolution logic.

In the example above, we were dealing with conflict resolution within a set, in particular the friends list associated with each User object. The merge operation that we built to handle conflict resolution is analogous to the resolution logic that is built into Riak sets. For more information on how you could potentially replace the client-side resolution that we implemented above, see our tutorial on Riak sets.