The Java SDK (in the comments) of ViewQuery indicates that the default Stale level is Update_After. In .NET, the default is indicated to be OK. I’d argue that the default should be OK as that’s the least intensive on the Couchbase cluster.
Thoughts?
The Java SDK (in the comments) of ViewQuery indicates that the default Stale level is Update_After. In .NET, the default is indicated to be OK. I’d argue that the default should be OK as that’s the least intensive on the Couchbase cluster.
Thoughts?
@unhuman actually, if you don’t specify any staleness level on the server it is UPDATE_AFTER, this is why the Java SDK indicates it is the default. So by default no staleness value is passed over to the server which then uses UPDATE_AFTER.
Where do you find that .NET always applies OK, even if you don’t set the staleness param explicitly?
Just working off of what the comment in the client says:
But, lower in the code, it seems to defer to whatever the server does by default:
Fundamentally, however, I think that querying views, by default, shouldn’t negatively affect server performance (circumventing the way views are re-indexed by the timer). If this is a server issue, can we just move the conversation to “Couchbase Server”?
Moved this thread to the Server category based on your comment…