WPS Portal project

Sunday, October 19, 2008

Alfresco Share : the new MS Sharepoint (and Quickr) killer ?

Last week I attend the Alfresco European Community conference in Munich Germany. There were some really exciting news about the Alfresco solution, and especially the sharepoint protocole for the Alfresco document management product, and also the new Alfresco SHARE collaborative solution.

I'm not familiar with Microsoft Sharepoint, but my understanding is that the Alfresco Share software is clearly provided as an alternative for the microsoft product.
SHARE offers most of the features that are now included in collaboration product (Tasks list, Calendar, Document management, blog, wiki, forum, etc). The Share platform uses of course the powerful Document Management options of the Alfresco repository (Share can be installed on top of Alfresco DM, or on a separate application server and connected remotely to the repository).

The share server runs Dashlet components (which are very similar to Portlets), and it is possible to extend the Share features by adding new custom Dashlets.

Moreover, Alfresco has now implemented the MS Sharepoint protocole and the Alfresco server can now be exposed a standard Sharepoint server. That means that it is considered as a standard Sharepoint server from the MS Office client (e.g in Word, you can connect to the Alfresco solution and manage the checkin-checkout on a document, exactly as if you were speaking to a Sharepoint server).
For the moment, all the sharepoint protocole is not implemented: for instance, there is still no way to connect Alfresco with the Outlook client...but I'm sure Alfresco will provide such feature very soon.

The goal of Share is to easily create and manage Sites (which are similar to small web site dedicated to a specific Project management). With a few clicks you can set-up a new Site and invite users to work on a new project.

So with the Share software, Alfresco now offers a collaborative solution, as well as a strong document management repository. It is clearly an alternative for MS Sharepoint customers.
I'm also sure that Share is also a big challenger for IBM Lotus Quickr :-)

Merging Websphere Portal and Quickr ?

According to IBM, there are still no plan to merge the WebSphere Portal and the Quickr product into a single software solution.
Moreover, most people say that they will remain two distinct softwares, because one main technical limitation is of course that a single JVM can hardly support both runtime.
So the roadmap is likely to manage these 2 solutions separately, but provide various facilities and option to implement a seamless integration for end-users.

For instance, integration can be based on REST services exposed by Quickr (and at the portal level, the RSS Portlet can be used).
Also the new webapp integrator tool seems to be a good way to "merge" the navigation of portal and quickr.
There are also some WP Portlet that can show the list of space the portal user is member of (simple link to access the remote Quickr server).

I found this article which describe how to integrate WCM (portal side) with Quickr:

Photo Book for IBM Lotus Quickr services for WebSphere Portal

The proposed approach is to consume the Quickr REST service from the WCM API, and to store link to Quickr data into WCM content.


Also, for those how think about running their WP Portlet directly on the Quickr server, you should know that even if your Portlet should run without any modification (assuming it is a JSR 168 Portlet), you will probably have to customize your Portlet code to transform it as a Quickr Business Component.
Basically, a Quickr Business Component is a Portlet which implements some Quickr java interfaces to manage the life cycle of the component and to ensure it will plug properly into the template application model of Quickr. For instance, by implementing these interfaces, you will ensure that the ACL applied on the composite application will be propagated also on all components (Portlets) included in the application.

Thursday, October 09, 2008

Limit with number of content items in WCM v5.x ? (Part 2)

Part 1

As you might know, I was in berlin for the WP conference and I got clear and also official informations about the wcm limit. I would like to share these info so that everybody could have the same background info:

- According to IBM, there is clearly a scalability issue with WCM v5,- Technically it is due to a limitation in the management of wcm indexes (which are partially stored in memory),
- The limit is not 'hardcoded' but deeply depends on the hardware configuration (especially memory),
- The limit which is observed usually for v5 is about 50.000 (but some customers have created more than that),
- If the limit is reached the impact is different depending on the server type:
- rendering server perf degrade slowly with the increasing number of items,
- however authoring server perf degradation is usually quicker and significant.

Note: even if some customer can manage more than 50.000 issues (sometimes up to 80.000 ?) that does not mean that your system will be able to support the same volume. It just shows that it is technically possible to do, but not that the system is still fully usable or offers acceptable performances.Also depending of the hardware configuration limit can vary significantly.

The only workaround solution in v5 seems to be to split of the authoring database in 2 parts (to reduce the number of items managed in each repository). FYI, we will work with the IBM support on this Authoring split scenario (please note that feasibility still has to be validated). But this should be considered as a plan B only as this would be a quite big effort for us to set-up (3 authoring server are required basically). So I hope the workaround solution will be provided by the support as a fix (assuming management of indexes could be improved in the application).

Websphere Portal conference Berlin 2008 / WP 6.1 feedback

This week I was in the WP conference in Berlin. Of course there are a lots of news I would like to share, but as I do not have the time yet, I just would like to share my overall feedback about the new 6.1 portal release.

First I would like to thank all IBMers I had the opportunity to met there, as they provided very valuable info and help me a lot. I really think that success of portal projects, as any highly technical and complexe project, mainly depends on people and competencies involved. So it is always very important to have the good resource or info at the right time. WP conference is a good way to gather such infos.


What about 6.1 ? Well to summarize I would say that:
- This is now a mature release (from the java standard implementation point of view),
- It is more open than ever (REST, new API, etc),
- Some really good new features (especially Client Side Aggregation),
- The roadmap of some features is not 'consistent' (or some bug/limitations exists),
- Migration effort from previous release 5.x is still very high.

------

Let me detail these topics:

>> Software design maturity:
- WCM data are now moved to JCR repository (the WCM has been completely re-designed),
- JSR 286 seems to be already implemented and supported (at least the major features like portlet cooperation).

All of these will allow customers to build custom solution on stable standards and without worrying about future migration, at least for a while.

---------

>> Plateform openess:
- A lots of portal features are now exposed through REST (atom) Web Services (or even new API). For instance you can access to the portal navigation (pages list) as a feed, or get the profil/membership of a user,
- Even WCM does offers REST facilities to manage content (CRUD operations are supported on content and sitearea).
- For some part of the code (e.g the WCM multi-language assets), IBM is also thinking about an open-source approach...

I'm sure this openess will really give the opportunity to extend existing features or to build new ones easily....that is clearly a good evolution.

----------

>> Cool new features:
- For me the major one is clearly CSA 'Client Side Aggregation' (assuming it is working well, this will clearly be a real improvement for navigation user-experience and page response time),
- Another good evolution is Remember me cookie (allow to identify the user without prompting for authentication, and so provide the ability to personalize some infos).
- WCM inline editing (6.0 feature).

These are at least the 3 main features we will surely integrate in our portal....but only once the 'as is' migration to 6.1 will be completed....

--------

>> Roadmap inconsistency, limitations:
- PDM does not exists in 6.1 anymore, but scripts to perform migration to 'Quickr PDM' will be provided only for Q1 2009 (reminder : end of support for v5 was sept 2008),
- Under IE, the CSA does not work for the WCM Rendering portlet (this is a bug that should be fixed soon),
- The role 'identified' (remember me cookie) is not implemented in WCM 6.1 (should be only in next version 7).

Well, I would just say that roadmap could have been better aligned....I know it is easy to say, but more difficult to do...

---------

>> Migration effort: you know my opinion about that. Even if migration scripts are supposed to be improved in 6.1 (and also better tested than for 6.0), the principle of side-by-side installation is still the same...this is still a MAJOR effort for customers to migrate to 6.1....

---------

In conclusion:
- 6.1 release is a major evolution (especially regarding the WCM software new technical design),
- Some evolutions are really an improvement for customer (but of course it depends on your project needs),
- Finally, I let you decide either or not the cost of migration is justified compare to these new benefits ;-)

Thursday, October 02, 2008

Limit with number of content items in WCM v5.x ?

As we currently experience crash of our portal production cluster, we have been in contact with IBM support. Of course, as we are still on v5.x, one of the first recommendation of the support was to migrate asap to v6.

But another feedback was that our crash issues (high cpu et frequent gc cycle) might also be due to some limit in the number of content items in WCM v5.x...
Indeed, our current total count of items is about 40.000, and some interlocutor at the support told us that some performances issues could be experienced with such a volume of data (either 40 or 50.000 items).
This was of course a very bad news for us, and a major risk for our portal...

On the other side, IBM tech sales expert told us that there is actually no pre-defined limit at the application level...however, as the WCM system stores a lot of data in memory (as caches, indexes, etc) the limit is the one of the hardware system: it is mainly the JVM memory that is the first bottleneck, but it could be also the CPU or the disk I/O...

So yes I think we can consider that there is a limit due to the design of the WCM application, even if this limit depends on the memory and hardware components.

The symptoms (if you think you might experience such problem) is supposed to be a slow degradation of system performances...So you will not see a sudden crash, but rather a chronic performance issue.

Another interesting information: IBM also told us that a few of their customers have been able to create and use WCM databases with more than 50.000 items (even 80.000 ?). So it is possible...

By the way, we still didn't found the cause of our portal crashes (but it is probably due to another problem than a system limit).
Another comment: since approx 2 month, our WCM authoring system is becoming more and more slow...IBM support is investigating and currently check if the problem is not related to WCM indexes....I hope we are not reaching the indexes limit :-)