Here is the ticket for this issue: https://issues.apache.org/jira/browse/IGNITE-2225
On Mon, Dec 21, 2015 at 10:00 PM, Vladimir Ozerov <[email protected]> wrote: > As per my I tickets, I have only single task targeting 1.5 - > https://issues.apache.org/jira/browse/IGNITE-2209 Hope it will be finished > in the nearest time. > > As per Externalizable issue, this could be due to my latests changes to > serialization logic. For now both Externalizables and objects with > writeObject/readObject methods are treated in the same way - they are > serialized using OptimizedMarshalling. This could be possible reason of > observed deserializations, but we need to double-check that. > > On Tue, Dec 22, 2015 at 4:34 AM, Dmitriy Setrakyan <[email protected]> > wrote: > > > Valentin, great catch! You should add this test to the cache suite and > make > > sure it fails, so we will be certain to fix it before the final release. > > Let’s also file a blocker ticket for it. > > > > Moreover, we should have an identical test for non-externalizable > classes, > > however I am not sure yet how to count the number of (de)serializations > in > > that case. Any ideas? > > > > D. > > > > On Mon, Dec 21, 2015 at 5:31 PM, Valentin Kulichenko < > > [email protected]> wrote: > > > > > Regarding IGNITE-2212. The original issue is reproduced only with > > optimized > > > marshaller, so I don't think it's critical. But at the same time I > > observe > > > strange behavior when I run the same test with binary marshaller: to > send > > > an Externalizable object across network, we serialize, deserialize, and > > > then serialize it again, all on client side :) > > > > > > I believe that's because our latest changes for Externalizables - > > > CacheObjectBinaryProcessorImpl.marshalToBinary() method doesn't > properly > > > handle them. Actually, I don't think we should call it in this case at > > all. > > > Alexey, can you take a look? I think this one should be fixed in 1.5. > > > > > > -Val > > > > > > On Mon, Dec 21, 2015 at 9:17 AM, Dmitriy Setrakyan < > > [email protected]> > > > wrote: > > > > > > > Alexey, > > > > > > > > About IGNITE-2212 - can you explain why the value is deserialized at > > all, > > > > even once, let alone multiple times? > > > > > > > > D. > > > > > > > > On Mon, Dec 21, 2015 at 8:57 AM, Alexey Goncharuk < > > > > [email protected]> wrote: > > > > > > > > > Status for the opened tickets from my side: > > > > > > > > > > https://issues.apache.org/jira/browse/IGNITE-2078The original > issue > > > > > reported has been fixed. Now sometimes the example returns > duplicate > > > > > records for TXT queries, this is related to changing topology not > > being > > > > > handled in TXT queries. This is an old issue and it is not critical > > for > > > > the > > > > > release > > > > > > > > > > IGNITE-2190 - Reported by Denis. Added test, fixed and merged to > > > > > igntie-1.5 > > > > > > > > > > The only issue that I had not time to take a look at is > > > > > https://issues.apache.org/jira/browse/IGNITE-2212- tx put triggers > > > > > multiple > > > > > deserializations of cache value. There is no a simple fix for this, > > > > however > > > > > I see that the fix version for this ticket is 1.6. Can somebody > > confirm > > > > > that it is not critical for the release? > > > > > On Dec 21, 2015 19:34, "Dmitriy Setrakyan" <[email protected]> > > > > wrote: > > > > > > > > > > > I would like to assemble a list of the remaining issues for the > 1.5 > > > > final > > > > > > release. Can everyone in the community please reply here with a > > list > > > of > > > > > > tickets you are still working on for 1.5 release? > > > > > > > > > > > > Thanks, > > > > > > D. > > > > > > > > > > > > > > > > > > > > >
