Re: [infinispan-dev] support for XA CacheStores in 6.0
On 11 Apr 2012, at 12:54, Manik Surtani wrote:
On 11 Apr 2012, at 12:37, Mircea Markus wrote:
> Not true. The recovery support we have implemented is there to solve just that and
become consistent. There's no "just as consistent" but only
"consistent" or "not consistent" :-)
Recovery assumes manual intervention though?
yes.
If so again you should document that cache stores only risk being
inconsistent if you are using recovery. Otherwise, the risk of inconsistency is the same
risk you have with a non-recovery-enabled in-memory container.
So basically the risk
of inconsistency when using cache loaders is there disregarding you using recovery or not.
I'll add a statement to further clarify that recovery won't be able to reconcile
the state.
Basically, you're trying to say Cache Stores don't support recovery. ;)
nor atomicity, as data can be partially applied to the store during commit.
Attachments:
- attachment.html (text/html — 2.2 KB)