each of CacheLoaderInterceptor, ReplicationInterceptor, and TransactionInterceptor
has a list of transactions for which they have been registered. after transaction
completion, the transactions are not removed from the list, so the lists grow without
bound. since these interceptors use
should the transaction be removed from the list at the end of
SynchronizationHandler.afterCompletion?
View the original post :
http://www.jboss.org/index.html?module=bbop=viewtopicp=3853136#3853136
Reply to the post :
http://www.jboss.org/index.html?module=bbop=postingmode=replyp=3853136
i'm using JBCv1.1 for the hibernate secondary-cache implementation. i'm getting a
catastrophic exception when a rollback occurs in certain situations. it is
catastrophic because locks are left locked and orphaned and the tree is generally in
an inconsistent state.
the problem is triggered by
this appears to occur because /foo/1 is listed in TransactionEntry.nodes twice.
when TreeCache.rollback processes the second /foo/1 node in TransactionEntry.nodes,
the call to findNode in _remove (re)creates the node for /foo/1 and adds it to
TransactionEntry.nodes (so now there are three in
getRegion does a similar test, should it be fixed as well?
instead of all the (repeated) string concatenations, would it be better to have
Region.getFqn() return a string that's always terminated by /?
-l
View the original post :
norbert wrote : it would be better design if Region.getFqn() returns an Fqn not a
String so we could use Fqn.isChildOf(Fqn) instead of String.startsWith()
|
| BtW: Fqn are not just concatenated Strings - In the longrun we have to find a way
to configure Regions for eviction based on
currently, Fqn.isChildOf is broken. if you attempt to check to see if a *parent* is
the child of a *child* you'll get an exception:
a = Fqn.fromString(/a)
| b = Fqn.fromString(/a/b)
| b.isChildOf(a)
| 1
| a.isChildOf(b)
| Traceback (innermost last):
| File console, line 1, in ?
bwang00 wrote : The CacheLoader in 1.1 release can persist the evicted data. So you
can retrieve from the backend store the next time.
|
when the CacheLoader reads the transient data from the data store, does it do a put
on the cache? that is, does it replicate the data to the rest of the
i've got some code i'd like to contribute (mods to LRUPolicy and an alternate to
LRUAlgorithm).
what is the stance on 1.4 based code?
what's the best way to contribute the code (via the forurm, is there an alternate
mailling list)?
thanks.
View the original post :
i was wondering why all the TreeCache methods throw Exception instead of something
more specific?
for example, setIsolationLevel does:
| if(tmp_level == null) {
| throw new Exception(TreeCache.setIsolationLevel(): level \ + level +
\ is invalid);
| }
|
why not
ok, great.
i've been looking at the 3.2 cvs branch. should i go ahead and move to the head? is
the head jboss-cache code compatible with the 3.2 server?
View the original post :
http://www.jboss.org/index.html?module=bbop=viewtopicp=3845946#3845946
Reply to the post :
there are two queues in a Region: nodeQueue and eventQueue. i understand that the
BoundedBuffer is thread-safe. however, since there are two queues that are being
operated on, it seems like there is a potential race in Region.putNodeEvent and
LRUAlgorithm.processQueues. is there a bit of
12 matches
Mail list logo