On Mon, Jun 26, 2017 at 2:31 AM, Yakov Zhdanov wrote:
> Guys, I see little inconsistency. Imagine user does 1000 puts with near
> cache enabled. Then user does some gets() and size() returns 1000 + N, more
> gets() and size() is 1000 + M. This is a bit weird. Can we have nearSize()
> on public AP
Guys, I see little inconsistency. Imagine user does 1000 puts with near
cache enabled. Then user does some gets() and size() returns 1000 + N, more
gets() and size() is 1000 + M. This is a bit weird. Can we have nearSize()
on public API? Any thoughts here?
As far as the original issue I would not
Hm... why not just return the key count? Can we really have nulls in the
near cache?
On Fri, Jun 23, 2017 at 8:25 PM, Mikhail Cherkasov
wrote:
> Hi all,
>
> GridCacheAdapter#size() has O(n) complexity and this can lead for delays
> during metric collection( https://issues.apache.org/
> jira/brow
