Mailing-List: contact user-help@ignite.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@ignite.apache.org
Date: Fri, 18 Aug 2017 14:22:12 -0700 (MST)
From: vkulichenko <valentin.kulichenko@gmail.com>
To: user@ignite.apache.org
Message-ID: <1503091332121-16303.post@n6.nabble.com>
In-Reply-To: <1503081565260-16298.post@n6.nabble.com>
References: <1503046298401-16278.post@n6.nabble.com> <1503074813309-16291.post@n6.nabble.com> <1503078096625-16294.post@n6.nabble.com> <1503081565260-16298.post@n6.nabble.com>
Subject: Re: 2.1 new feature - warm up the cache / memory
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
archived-at: Fri, 18 Aug 2017 21:22:16 -0000

Marco,

It's not a background warmup, there is no explicit warm-up at all. Ignite
starting with 2.0 uses page memory architecture, i.e. data is organized in
pages. Any page can be stored both in memory and on disk, and access to this
page is transparent to client, regardless of where it resides. This means
that when you access it (run a SQL query, for example), Ignite will detect
if it's in memory or not, and fetch into memory if it isn't. When if
everything is on disk (which is the case after cluster restart), you can
still access any data.

From architecture standpoint, there is indeed a lot in common with
databases. However Ignite follows memory-centric approach rather than
disk-centric, plus it's distributed and scalable.

Makes sense?

-Val


--
View this message in context: http://apache-ignite-users.70518.x6.nabble.com/2-1-new-feature-warm-up-the-cache-memory-tp16278p16303.html
Sent from the Apache Ignite Users mailing list archive at Nabble.com.