Im affraid i said it all in the title.

After some nodes created and ran the cron the search engine rewrite the url to the default domain, even if the node are published only in other domains, I mean in the "publishing option" the node are checked for another domain, and the "sourse domain" are set for for a non-default domain.

I read previus issues and try the garland theme and try to use the Domain Source but it still dosn´t work.

By the way i use a separated alias_url table for each domain using the "Table prefixing" settings of the module, to allow similar url in diferents domain, ex: http://gescon.cav.sld.cu/inicio and http://caed.cav.sld.cu/inicio

Any ideas please.

PD: Excuse my english please

CommentFileSizeAuthor
#5 419012-search.patch1.08 KBagentrickard

Comments

neopulse’s picture

Im sorry it doen´t rewrite the url to the default domain, it does to the current domain, I mean if im in the http://gescon.cav.sld.cu and made a search the search result redirect al url to http://gescon.cav.sld.cu/nodesresult even if these nodes are content only for another domain.

agentrickard’s picture

Category: bug » support

Please read section 4.3.1 of the provided documentation. And remember that users with the administer nodes permission can always see all content.

neopulse’s picture

Sorry, i already did (read the 4.3.1 section), and i try every single option, after that rebuild the index search and run the cron but the trouble remains.

If i try from another user, im still redirected to http://primarydomain/node/nodenumber, and trying to access to this page: "access denied" page appear.

any ideas

pd: excuse my english please

agentrickard’s picture

You have either missed a step or there is a server problem somewhere.

Are you loading the include file correctly in settings.php?

Are you using any other implementations of custom_url_rewrite_outbound()?

Did you accidentally remove the node/%n pattern from the "Node link patterns" setting? If so, then the links cannot be identified as nodes.

It may also be possible that the path aliases are a problem. Take a look at #300454: Inter domain linking and see if that helps.

agentrickard’s picture

Category: support » bug
Status: Active » Needs review
StatusFileSize
new1.08 KB

I found a case where something like this happens if a source domain is not set for specific nodes.

Does this patch help?

agentrickard’s picture

Assigned: neopulse » Unassigned

I am seeing this behavior more clearly now. Will dive in tomorrow. It looks like we broke something....

agentrickard’s picture

Status: Needs review » Needs work

I think I may see the problem, and it is a deliberate choice.

If you create content before enabling Domain Source, content is not necessarily assigned to a source domain. This is by choice. The assumption is that you will use the Domain Content module (or node editing) to update the source domains. This is documented:

----
2.2 Warning

When this module is enabled, existing content is not automatically assigned
to a source domain.

You must either edit the nodes individually or use the Domain Content
module to perform batch edits.

However, if Domain Source is turned on, nodes without a specified source domain can link to the wrong site (typically the default site), which is not the preferred behavior.

I am open to suggestions about the best way to handle this. One option is to simply add records to Domain Source when the module is enabled.

neopulse’s picture

I'm afraid this is not my case, my first step was install the module, then i create the domains and my last work was create the nodes.

I fallow every think as the install say but nothing, my only change was the url_alias prefix for every domain.

neopulse’s picture

I'm not using using any other implementations of custom_url_rewrite_outbound(), but the fckeditor module modify my "Node link patterns" setting, it put the paragraph label int the pattern box, i removed and exclude this page for loading the fckeditor again throw its own administration page, now i'm afraid the node link patterns are corrupted.

neopulse’s picture

Yeahhhhhh, i solved, with your help of course. Actually I should said WE SOLVED

As i said before the fckeditor module put "his hand" on my Node link patterns, you gave my the light remember?? Here.

After avoid the fckeditor load in the domain option page every things seems ok.....but not. I remove the paragraphs labels around the links patterns but they were still corrupted by...... spaces. I meant at the end of every link patterns (one per line) the were some void spaces or white spaces, i don't know how to named in English, just like the space bar pressed two or three times at the end of every line, and this corrupted the links patterns.

This was really close to drive me crazy for a week, and i almost drive crazy, to every one in the drupal site who fallow the project domains issues, specially to agentrickard who are alway here, to helps us.

Thank very much my friend and excuse my English, please

neopulse’s picture

neopulse’s picture

Status: Needs work » Closed (fixed)