Mailing-List: contact user-help@ignite.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@ignite.apache.org
Date: Thu, 2 Jun 2016 13:31:37 -0700 (PDT)
From: bintisepaha <binti.sepaha@tudor.com>
To: user@ignite.apache.org
Message-ID: <1464899497196-5386.post@n6.nabble.com>
Subject: Cache.put delays
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
archived-at: Thu, 02 Jun 2016 20:47:15 -0000

We are using ignite 1.5.0-final version. While testing on a single server
node we see this behavior with cache.put(). We know we have just loaded an
object in the cache from the DB Store, as soon as that is put in the cache,
the writebehind is also kicked off. But right after loading it when we do a
cache.get for the same key the object returned is null. I was debugging this
code and I had a breakpoint after cache.get() line and a few seconds later
when I inspected the same cache.get() call, I saw the object being returned
correctly in the eclipse variable window.

As soon as I wrap the loading of this object from the DB store in a
transaction, I don't see this issue at all. We do that we are invoking load
sequentially for each object and the object is not loaded twice. SO don't
see a locking issue.

Is this the same as issue described here?
https://issues.apache.org/jira/browse/IGNITE-2407?jql=project%20%3D%20IGNITE%20AND%20fixVersion%20%3D%201.6%20ORDER%20BY%20updated%20DESC%2C%20priority%20DESC%2C%20created%20ASC

However, its important to note that we see this issue with a single server
node, so not sure why partitioned caches might be an issue here.

We are seeing more issues with cache.put and get. the behavior is random but
the results are as if all puts are not seen by gets. Any help is
appreciated.

Thanks,
Binti


--
View this message in context: http://apache-ignite-users.70518.x6.nabble.com/Cache-put-delays-tp5386.html
Sent from the Apache Ignite Users mailing list archive at Nabble.com.