Livestore is a transparent JDBC distributed data cache, which holds data local to the application server, reducing common performance bottlenecks associated with database access. Versoi 1.1 adds a app behaviour based self-configuraiton, and a web monitoring console that allows live cache configuration changes and cache status checking.

Check out Isocra Livestore.

Press Release
----------------
Cambridge, UK, 3rd July, 2003; Isocra announces the release of the latest version of its flagship product, livestore. livestore is a transparent JDBC data cache, which holds data local to the application server and provides a dramatic performance improvement for many users of J2EE systems. The latest release of livestore builds on this basis adding significant ease-of-use features.

Unlike other caching solutions, livestore provides code free integration which means that applications do not have to be specially adapted to take advantage of the locally held data. The product can be slotted into an existing or new application by simply pointing the application at livestore instead of at the existing JDBC driver. None of the security and integrity features of the application are compromised and, because of its J2EE standards compliance, there is no technology lock-in.

Benchmarking tests show speed gains in excess of ten and 100-fold for database queries. This increased application performance, together with reduced network and database loading, results in important cost and time savings over re-engineering or purchasing larger application and database tier hardware.

livestore 1.1 adds a built-in configuration and monitoring capability, the livestore console, allowing users to inspect and modify the state of a running livestore installation. The livestore console is available to users via a Web interface and is easily usable across a user's network requiring only the availability of a web browser. The console provides easy access to:

  configuration information so that livestore users can inspect and modify the configuration of a running livestore installation using a simple web-based interface,

  cache status so that a user can easily see those queries that are accessing the cache and any that might be causing a cache flush.
 
In order to make using the product even easier livestore 1.1 provides for automatic configuration. In this mode livestore will automatically add tables and indexes to its configuration as it sees a need for them based on the SQL that is actually being used. In this way, the user can start with an empty configuration and let livestore just configure itself. The console can be used at any time to make the configuration persistent, avoiding the requirement for livestore to re-configure itself the next time the application is run.

To try livestore, download a free trial copy from the Isocra website on www.isocra.com/livestore.

For more information mail livestore-sales at isocra dot com or call +44 1223 551610.

So, let say that now I have to forget everything I've learn using things like Hibernate?

I'm really not sure about all this. I'm getting CRAZY over here! Every day, a new technology comes up and drive me nuts!

If livestore is really good, why use Hibernate, Toplink, and other cool stuffs?

As far as I can tell, livestore is caching/passthrough JDBC driver replacement. It doesn't replace Hibernate or Toplink as they are not JDBC drivers, they are O/R mappers.

If you can drop in LiveStore and it improves your performance, why wouldn't you do it?

Peace,

Cameron Purdy
Tangosol, Inc.
Coherence: Easily share live data across a cluster!

Yes, it can improve performance very much, but I do not think it can work for all of use cases. View definitions, triggers, UDF, updates done by scripts are not known for JDBC driver and it can not implement 100% safe cache
(Is it documented ?).


It works for trivial "readonly" style web applications, but I think it is better to cache content for this kind of applications.
It can be "transparent" too, but much more trivial and "fast":
Cache GET methods and clear all after POST in Filter.

Sorry, if I am too sceptic.

From what I've seen, you can use it even with stored procedures, triggers, etc. by configuring it to understand potential dependencies. However, obviously the cache itself will lose some effectiveness in those cases. I've always liked the LiveStore idea, though, because they focus on making customers' lives simpler (in this case by being able to "drop in" a caching JDBC driver).

Peace,

Cameron Purdy
Tangosol, Inc.
Coherence: Easily share live data across a cluster!

How are select for update statements handled at the cache level?

Does Coherence Cache, which they use ;-), provide concurrency/locking mechanisms here to mimic this. Will the transactional semantics be the same? Just curious.

I need to think more about the products feature set. Initially this seems very useful.

Regards,

William Louth
"Test and Tune with Insight"
www.jinspired.com

Do you have any figures for an application other then the one described very briefly on the website alongside performance data showing alot of savings in time.

I was thinking along the lines of ECperf of SPECJAppserver2002. Maybe providing summary reports with and without your caching solution in a non-clustered and clustered environment.

Regards,

William Louth
"Test and Tune with Insight"
www.jinspired.com

William,

If you download the product then there is an included document which goes into various scenarios, including the one mentioned on the website. To be honest it's a bit difficult because we need to avoid talking about specific appservers and DBMSs.

The performance available with the product is, not surprisingly, highly dependent on what the application is. One customer, for example, got a 100-fold improvement in performance. Probably as a result of a poor DBMS/JDBC installation though.

If you want some more detailed information, then feel free to contact me directly.

Regards,

Tim Hoverd
www.isocra.com
tim.hoverd@isocra.com

livestore is definitely not aimed at trivial read-only applications. It's a full distributed write-through cache. So if you have multiple app-servers all using the livestore cache then when one of them writes data, all the caches know about it.

Obviously, if there's an external application that is updating the database, then you need to let the cache(s) know, but there is an API to do this.

...

On the Toplink et al front, as someone has already pointed out Toplink is an O/R mapping tool as well as a cache so livestore isn't a direct replacement. But if you're using CMP+livestore, then livestore provides the caching transparently and you get all the benefits and none of the hassle of having to hand-roll your persistence layer.

See
http://www.mysql.com/doc/en/Query_Cache.html,
this must be a simple and safe way for the most of use cases too.
I have used cache in home made driver a few years ago, but I have found a lot of better ways later.