#1.
Couch-based test is
That only one thread is asyncGet,
Only one thread to the set, and
Indicates the number of times the behavior.
In the case of a typical DB select the insert you see results much faster than the process,
In contrast, the two set couchbase faster than asyncGet shows.
If a simple comparison of the results of this more than 100 times faster set’m representing.
Non-synchronized manner using non-synchronized manner even when the function of the asyncGet
I wonder why the slower set.
#2.
Couchbase OS and the server is the same for all preferences.
Why only 1 server poisonous toxic ram usage is higher and …
In the case of 1 server ram the Full DISK swap two having been finally determined that this will continue to happen.
Why this happens, and
DISK swap is recommended to prevent this happening in the OS area wondering how much memory area size.