You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@archiva.apache.org by "Robert Simmons jr. Simmons" <rs...@icat.com> on 2007/05/10 18:17:18 UTC

Issues with Archiva from a User

Greetings, 

I have been using Archiva at work, in a professional development
environment, now for a couple of months and I wanted to talk about some of
the issues that I have with it to bring them up for discussion among the
developers. Hopefully out of this will result some feature requests or bug
fixes. The issues I have will be in the order of importance to me and my
users: 

1) No support of LDAP authentication. This has got to be the highest issue
for me. My company has a single sign-on LDAP server and I am not able to
configure Archiva to use that server in order to do sign ons and user role
verifications. What I think should happen is that the user should be able to
configure the web application WAR like any other web app using the J2EE JAAS
architecture. That way we could use an embedded properties file, another
service or LDAP. 

2) Deployment on Tomcat. In a corporate environment where Tomcat is
standard, you simply can't tell network admin you will be running plexus. I
have tried many times to deploy it on the most recent tomcat, all with
dismal results. The deployment of the web application should be application
server agnostic as much as possible with the application server specific
deployment descriptors for several application servers included in the war.
Supported platforms out of the box should be Tomcat, Jetty, JBoss, WebLogic,
and WebSphere.

3) Transparent location of manually deployed artifacts. If I just copy
artifacts into my repository, the search and indexing functionality doesn't
find them properly.

4) Abhorrent documentation when it comes to setting up the ability to deploy
to the repository with 'mvn deploy'. Its a pain to set up and get just right
with the webdav URLs and it seems like it has changed several times since
wiki pages were created. There needs to be a document that has the process
and setup once and for all and that process needs to be finalized so that
there is no more lack of stability in the UI.

5) Ability to search on remote archives. First of all I don't get how the
search works. I want to be able to enter "hibernate" and have it show me all
the hibernate artifacts it has. Furthermore, I would like to have the
ability to index the remote archives and do a search and be informed that
the artifact is in a remote archive and allow you to see the dependency POM
snippet.

In conclusion, I would love to help but I don't know the plexus container or
the stuff surrounding plexus so I am a bit of a loss as to what I could do.
If these things were worked out then archiva would truly be ready for heavy
development environments.


Robert Simmons jr. 
Senior Software Engineer
ICAT Managers, LLC 
720-271-9343
rsimmons@icat.com 


 
Confidentiality Note: This message contains information that may be confidential and/or privileged. If you are not the intended recipient, you should not use, copy, disclose, distribute or take any action based on this message. If you have received this message in error, please advise the sender immediately by reply email and delete this message. Although ICAT Managers, LLC scans e-mail and attachments for viruses, it does not guarantee that either are virus-free and accepts no liability for any damage sustained as a result of viruses.  Thank you.


Re: Issues with Archiva from a User

Posted by Joakim Erdfelt <jo...@erdfelt.com>.
Robert Simmons jr. Simmons wrote:
> Greetings, 
>
> I have been using Archiva at work, in a professional development
> environment, now for a couple of months and I wanted to talk about some of
> the issues that I have with it to bring them up for discussion among the
> developers. Hopefully out of this will result some feature requests or bug
> fixes. The issues I have will be in the order of importance to me and my
> users: 
>   
We welcome all suggestions. (we'll even put most of them on the roadmap!)
> 1) No support of LDAP authentication. This has got to be the highest issue
> for me. My company has a single sign-on LDAP server and I am not able to
> configure Archiva to use that server in order to do sign ons and user role
> verifications. What I think should happen is that the user should be able to
> configure the web application WAR like any other web app using the J2EE JAAS
> architecture. That way we could use an embedded properties file, another
> service or LDAP. 
>   
LDAP support is something our chosen security framework, plexus redback, 
has slated for support.  We chose not to use JAAS as it was not flexible 
enough for our needs.
Stay tuned, it will arrive, but not as a commit on archiva.
> 2) Deployment on Tomcat. In a corporate environment where Tomcat is
> standard, you simply can't tell network admin you will be running plexus. I
> have tried many times to deploy it on the most recent tomcat, all with
> dismal results. The deployment of the web application should be application
> server agnostic as much as possible with the application server specific
> deployment descriptors for several application servers included in the war.
> Supported platforms out of the box should be Tomcat, Jetty, JBoss, WebLogic,
> and WebSphere.
>   
We have many reported problems of tomcat deployment issues, but no details.
We test on Jetty, in fact, the plexus standalone is just kicking off 
jetty and running the war, no magic (honest!).
We have also reports that WebLogic and WebSphere work with no real 
issues (apart from setting up the JNDI resources for the Databases and 
Java Mail)

Have you tried these docs ?
http://docs.codehaus.org/display/MAVENUSER/Archiva+on+Tomcat

If you could detail the setup you have attempted on 
http://jira.codehaus.org/browse/MRM-333 that would be of great help to 
isolate the conditions that are causing all of the mysterious problems 
people have been reporting.

> 3) Transparent location of manually deployed artifacts. If I just copy
> artifacts into my repository, the search and indexing functionality doesn't
> find them properly.
This is the Repository Scanning angle.  In the original code-tree (archiva 0.9) this detection and indexing would often break.  It has been completely overhauled with the new trunk (1.0-alpha-1).  

We just finished off the new administration screens and have a host of bugs to fix around them.  This shouldn't take more than a few days.  Stay tuned in this mailing list for a declaration of 'initial stability' of the trunk.

> 4) Abhorrent documentation when it comes to setting up the ability to deploy
> to the repository with 'mvn deploy'. Its a pain to set up and get just right
> with the webdav URLs and it seems like it has changed several times since
> wiki pages were created. There needs to be a document that has the process
> and setup once and for all and that process needs to be finalized so that
> there is no more lack of stability in the UI.
>   
This is a pet peeve of mine.
This is a webapp!  Put the documentation online, in the war, live! ( as 
well as on http://maven.apache.org/archiva/ )
Our priorities at this very moment is a stable trunk.
Once the trunk stabilizes again, documentation should be able to be 
created easier.
> 5) Ability to search on remote archives. First of all I don't get how the
> search works. I want to be able to enter "hibernate" and have it show me all
> the hibernate artifacts it has. Furthermore, I would like to have the
> ability to index the remote archives and do a search and be informed that
> the artifact is in a remote archive and allow you to see the dependency POM
> snippet.
Remote archive indexing is a nightmare of bandwidth for the remote repositories.
As it stands, the central repository uses terabytes of bandwidth a week.
This request isn't exclusive for archiva, even the IDE plugin developers have asked for this feature.
We are working on establishing a set of services that a remote repository can provide that other archiva instances can make use of.  Static resources, downloadable indexes, even web services.

> In conclusion, I would love to help but I don't know the plexus container or
> the stuff surrounding plexus so I am a bit of a loss as to what I could do.
> If these things were worked out then archiva would truly be ready for heavy
> development environments.
>   
Deep knowledge of Plexus isn't a requirement to help with archiva, but 
it does help.
Stay tuned for a "Hacking Archiva" set of documentation once we stablize 
the majority of archiva trunk.
> Robert Simmons jr. 
> Senior Software Engineer
> ICAT Managers, LLC 
> 720-271-9343
> rsimmons@icat.com 
>   
- Joakim Erdfelt
  joakime@apache.org
  maven archiva commitor