It all depends on you intent for the value. Are you looking for unique identifiers, then GUID is best. If you are looking for sequential ordering, then snowflake. If you need counters, then AtomicInteger like you were asking.
May you could provide a bit more context on the intended use. -Jake On Mon, Jul 17, 2017 at 9:34 AM Amit Pandey <[email protected]> wrote: > I think block allocation is something I am more interested in. As that > should give better performance also. > > Regards > > On Mon, Jul 17, 2017 at 10:01 PM, Anthony Baker <[email protected]> wrote: > >> You can build concurrency primitives like AtomicInteger using the >> DistributedLockService. However, cluster-wide synchronization is really an >> anti-pattern in distributed computing and should be used sparingly. >> Ideally, you should find algorithmic alternatives such as block leasing >> with local allocation, twitter snowflake, etc. >> >> Anthony >> >> >> > On Jul 16, 2017, at 8:15 AM, Amit Pandey <[email protected]> >> wrote: >> > >> > Hi Guys, >> > >> > We had a hazelcast app we want to port to geode. There are some usages >> of AtomicLOng of hazelcast which we need to port, is there a best >> practice/alternative to achieve the same in Apache Geode? >> > >> > regards >> >> >
