Jens, that is a great thought. Lifecycle management would be amazing and I
think we can definitely think about that. Right now, my thought of the services
that are complimentary to the Cache. i.e the Cache uses them, but they
themselves are not dependent on them.
I still see things like HttpSer
One aspect this conversation has brought to mind is whether these services will
require some kind of lifecycle management? If any of them require an instance
of a Cache (and perhaps other components) they will need to be aware of the
cache restarting (happens during reconnect).
Is that going to
Hi there Dan,
We (Patrick and I) are in violent agreement about adding new singletons to the
product. This singleton is merely there to avoid the complete utter re-write of
the system to wire in said ServiceRegistry into the cache.
Yes, we are merely replacing 1 singleton with another, but we s
Hi there Jens,
No this is not intended to replace the Service loading mechanism that is
currently there.
This is merely a registry that is to aid the looking up of services (like
ClassLoaderService or MetricService) within the code without having to
explicitly add that service onto every const
A word of caution here...
I'd like to see us start moving away from "internal" APIs even, as much as
possible. Moving away from Singletons is no-brainer, but less obvious, is
moving things to proper public APIs and SPIs so that frameworks and tooling can
extend Geode in interesting ways. SDG
It might be better to hang things off InternalCache rather than create a new
singleton. The cache is already a singleton, but we've been working on removing
that by plumbing the cache everywhere. The cache is effectively our context
object, and once we've finished removing calls to Cache.getInst
Hi Jakov,
Do you see this issue if you drain the worker node before shutting it down? For
reference:
https://kubernetes.io/docs/tasks/administer-cluster/safely-drain-node/. I know
this doesn’t directly answer your questions, but it seems like draining the
node first would help the members rece
Hi Ernie,
Thank you really much for trying to help.
Unfortunately we haven't been able to reproduce this issue without k8s.
I would like to share some additional info about this issue. We know
that this is not the "correct" way to shut down member gracefully, since
it's complete TCP communica
Hi Udo,
Is the intention of this to replace (and extend) what we're doing with the
traditional service loading mechanism today? i.e. how we're loading everything
that extends CacheService (for example HttpService, LuceneService,
GeodeRedisService, etc.).
Thanks
--jens
On 10/15/20, 7:25 PM,