Hi,
My use case is to collocate the data and execute the task on that node. But
I want to execute >1000 number of such tasks in collocate task mode and all
such tasks should be executed in a single transaction. If any one of the
task fails, I want to rollback the complete transaction. The reason to
execute the task collocate mode is my input data is going to be huge. Can
you please explain how to implement behavior using your solution?
As per your solution I will be executing transaction.commit 1000 times if
I have 1000 such compute tasks but if any one of it fails I do not have a
way to rollback complete transaction.
Is there any alternative to implement this?
Thanks,
Prasad
On Mon, Feb 12, 2018 at 6:22 PM, Ilya Lantukh <ilant...@gridgain.com> wrote:
> Hi,
>
> The fact that code from invoke(...) is executed on node that initiated
> transaction ("near node" in ignite terminology) is a known issue. There is
> a ticket for it (https://issues.apache.org/jira/browse/IGNITE-3471), but
> it hasn't been fixed yet.
>
> To solve your initial goal, you might want to start transaction on the
> primary node for your key. It can be achieved by using
> ignite.compute().affinityRun(...), but in this case you have to start
> transaction inside affinityRun closure.
>
> Like this:
> ignite.compute().affinityRun(cacheName, key,
> () -> {
> try (Transaction tx =
> Ignition.ignite().transactions().txStart(...))
> {
> cache.invoke(key, entryProcessor);
>
> tx.commit();
> }
> }
> );
> }
>
> In this case you will minimize overhead to modify entry in cache -
> entryProcessor will be executed only on nodes that own the key, and stored
> value shouldn't be transferred between nodes at all.
>
> Hope this helps.
>
>
>
> On Mon, Feb 12, 2018 at 1:57 PM, Prasad Bhalerao <
> prasadbhalerao1...@gmail.com> wrote:
>
>> Hi,
>>
>> I am trying to test the distributed transaction support using following
>> piece of code. While debugging the code I observed that code executes on
>> client node first and after doing commit the code executes on a node which
>> owns that kay.
>>
>> What I am trying to do is, to collocate the data to avoid the network
>> call as my data in real use case is going to big. But while debugging the
>> code, I observed that entry processor first executes on client node, gets
>> all the data executes the task. and after commit executes the same code on
>> remote node.
>>
>> Can someone please explain this behavior? My use case to execute the task
>> on nodes which owns the data in single transaction.
>>
>> private static void executeEntryProcessorTransaction(IgniteCache<Long,
>> Person> cache) {
>> Person val=null;
>> try (Transaction tx = Ignition.ignite().transactions
>> ().txStart(TransactionConcurrency.OPTIMISTIC,TransactionIsolation.SERIALIZABLE))
>> {
>> long myid =6l;
>> CacheEntryProcessor entryProcessor = new MyEntryProcessor();
>> cache.invoke(myid, entryProcessor);
>> System.out.println("Overwrote old value: " + val);
>> val = cache.get(myid);
>> System.out.println("Read value: " + val);
>>
>> tx.commit();
>> System.out.println("Read value after commit: " +
>> cache.get(myid));
>> }
>> }
>>
>>
>>
>> Thanks,
>> Prasad
>>
>
>
>
> --
> Best regards,
> Ilya
>
