Hi,

in a (quite large and turnover significant) project one of the teams has 
concluded they need per-datacenter singleton semantics to create strictly 
monotonous timestamps in the microsecond area[1].

The team came up with the idea to use Akka cluster singletons to assure only 
one clock service is active at any given point in time.

Set aside the suitability of the solution in general, I really wonder whether 
Akka cluster singletons are yet ready and stable (roadmap wise) enough to put 
them at the heart of a substantial part of that software system.

Do you (the Akka developers) see cluster singletons as being truly production 
ready for such a scenario or would you suggest to stick to other leader 
elections tools such as etcd or Zookeeper for the time being?

Especially since it is rather new and consensus algorithm implementations do 
not really have a track record of being easy to get right in the first 
iterations. I also read the manual vs. automatic restart section of the docs 
that makes me wonder how all the complexity will affect production quality of 
what we are building here.

Unavailability periods of the clock service is acceptable in the range of seconds, 
incorrect timestamp behaviour (accidentally running >1 singleton) is not 
acceptable at all as it would potentially screw up significant business data.

It would be a great help if you could comment on using Akka cluster singletons 
in that context.

Jan


[1] For explicitly ordering Cassandra mutations in case you wonder

--
     Read the docs: http://akka.io/docs/
     Check the FAQ: http://doc.akka.io/docs/akka/current/additional/faq.html
     Search the archives: https://groups.google.com/group/akka-user
--- You received this message because you are subscribed to the Google Groups "Akka User List" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/akka-user.
For more options, visit https://groups.google.com/d/optout.

Reply via email to