Google Cloud Platform Blog
Product updates, customer stories, and tips and tricks on Google Cloud Platform
Best practices for App Engine: memcache and eventual vs. strong consistency
Tuesday, December 17, 2013
We have published two new articles about best practices for App Engine. Are you aware of the best ways to keep Memcache and Datastore in sync? The article
Best Practices for App Engine Memcache
discusses concurrency, performanceand migration with Memcache to make you aware of potential pitfalls and to help you build more robust code.
Do you know how to make your App Engine application faster and more scalable by using eventual consistency? If not, take a look at a
new article
that explains the difference between eventual and strong consistency. The paper will help you leverage Datastore to scale your apps to millions of happy customers.
Concurrency, performance and migration in memcache
Memcache is a cache service for App Engine applications that is shared by multiple frontend instances and requests. It provides in-memory, temporary storage that is intended primarily as a cache for rapid retrieval of data that's backed by some form of persistent storage, such as Google Cloud Datastore.
Using Memcache will speed up your application's response to requests and reduce hits to the datastore (which in turn saves you money). However, keeping Memcache data synchronized with data in the persistent storage can be challenging when multiple clients modify the application data.
Transactional data sources, such as relational databases or Google Cloud Datastore, coordinate concurrent access by multiple clients. However, Memcache is not transactional, and there's a chance that two clients will simultaneously modify the same piece of data in Memcache. As a result, the data stored may be incorrect. Concurrency problems can be hard to detect because often they do not appear until the application is under load from many users.
With App Engine, you can use the “compare and set” (Client.cas()) function to coordinate concurrent access to memcache. However, if your application uses the compare and set function, it must be prepared to do the error handling and retry.
We recommend that you use the atomic Memcache functions where possible, including incr() and decr(), and use the cas() function for coordinating concurrent access. Use the Python NDB API if the application uses Memcache as a way to optimize reading and writing to Google Cloud Datastore. Read more
Best Practices for App Engine Memcache
in our newly published paper.
Balancing strong and eventual consistency
Web applications that require high-scalability often use NoSQL which offers eventual consistency for improved scalability. However, if you're used to the strong consistency that relational databases offer, it can be a bit of a mind shift to get your head around the eventual consistency of NoSQL data stores. Google Datastore allows you to choose between strong and eventual consistency, balancing the strengths of each.
Traditional relational databases provide strong consistency of their data, also called immediate consistency. This means that data viewed immediately after an update will be consistent for all observers of the entity. Use cases that require strong consistency include knowing “whether or not a user finished the billing process” or “the number of points a game player earned during a battle session.” It also means that all requests to view the updated data are blocked until all the writes required for strong consistency have finished.
Eventual consistency, on the other hand, means that all reads of the entity will eventually return the last updated value but might return inconsistent views of the data in the meantime. For example, knowing “who in your buddy list is online” or “how many users have +1’d your post” are cases where strong consistency is not required. Your application can get higher scalability and performance by leveraging eventual consistency, because your application won't have to wait for all the writes to complete before returning results.
The following two diagrams illustrate strong versus eventual consistency:
Eventual conistency
Strong consistency
To learn more about the differences between eventual and strong consistency and to learn how to take advantage of each read our
article
on the technical solutions portal at
cloud.google.com/resources
.
-Posted by Alex Amies, Cloud Solutions Technical Account Manager
Free Trial
GCP Blogs
Big Data & Machine Learning
Kubernetes
GCP Japan Blog
Firebase Blog
Apigee Blog
Popular Posts
Understanding Cloud Pricing
World's largest event dataset now publicly available in BigQuery
A look inside Google’s Data Center Networks
Enter the Andromeda zone - Google Cloud Platform’s latest networking stack
New in Google Cloud Storage: auto-delete, regional buckets and faster uploads
Labels
Announcements
193
Big Data & Machine Learning
134
Compute
271
Containers & Kubernetes
92
CRE
27
Customers
107
Developer Tools & Insights
151
Events
38
Infrastructure
44
Management Tools
87
Networking
43
Open
1
Open Source
135
Partners
102
Pricing
28
Security & Identity
85
Solutions
24
Stackdriver
24
Storage & Databases
164
Weekly Roundups
20
Feed
Subscribe by email
Demonstrate your proficiency to design, build and manage solutions on Google Cloud Platform.
Learn More
Technical questions? Check us out on
Stack Overflow
.
Subscribe to
our monthly newsletter
.
Google
on
Follow @googlecloud
Follow
Follow
Follow