4 ms·
This is extremely important to think about if you're building on top of cloud services. Misunderstanding consistency models and limitations can make or break y
by cle 10y ago
This is extremely important to think about if you're building on top of cloud services. Misunderstanding consistency models and limitations can make or break your application and cause serious data quality problems.
This seems relatively misleading, at least for S3. S3 is similar to the comments for Google Cloud. List operations are generally eventually consistent. RAW is strongly consistent. Read-after-overwrite is eventually consistent.
I can only comment on S3 because I have a lot of experience with it.
- jamesblonde 10y ago"Read-after-overwrite is eventually consistent." This means update to normal people.
- cle 10y agoRead-after-overwrite makes it clear that it's in contrast to and distinct from read-after-write. Sometimes different words mean the same thing, and you pick the word based on its context. And in this context, the readers aren't "normal people"--whatever that means--they're engineers. Where the specific meaning of "update" and "write" can vary depending on an even more specific context.
- dastbe 10y agogoogle cloud storage is strongly consistent on both initial and update writes, as well as deletes. It would be better if they compared object storage based on request types than saying "Eventual, except for..." vs. "Mixed". That tells me nothing.