Task #10273 (closed)
Bug: webstart urls hardcoded
Reported by: | wmoore | Owned by: | atarkowska |
---|---|---|---|
Priority: | critical | Milestone: | OMERO-4.4.9 |
Component: | Web | Version: | 4.4.8 |
Keywords: | n.a. | Cc: | cxallan, jburel, jamoore, atarkowska, m.woodbridge@…, julien.jourde@… |
Resources: | n.a. | Referenced By: | n.a. |
References: | n.a. | Remaining Time: | n.a. |
Sprint: | Blocker 4.4.9 (1) |
Description
From ome-users:
On 01.02.2013 15:09, Julien Jourde wrote:
Hi Omero-users,
I just installed a 4.4.5 omero server in my lab. Everything's fine
except when I try to launch Insight using the jnlp protocol.
The problem seems to come from the redirection of the port 80 to the
port 443 in nginx.
I'm using "rewrite https://$server_name$request_uri? permanent;" in my
omero.conf file to accomplish that. But the jnlp file can't deal with
that. It contains urls like
"http://my.omero.server.fr/webstart/jars/insight.jnlp" where it shoud be
"https://my....". Launching the jnlp leads to a launchfile error because
it tries to charge a web page unproperly created by nginx (see below).
Did I made something wrong by using redirection rules ? Is there a way
to do it more properly ?
On 01.02.2013 15:12, Niko Ehrenfeuchter wrote:
apparently we do have an exception for webstart in the rewrite rules:
if ( $uri !~ /webstart/jars/insight.jnlp ) {
rewrite https://$server_name$request_uri? last;
}
This workaround seems to get things going here, but anyway I think the hardcoded "http" in the jnlp file should rather get fixed.
Cheers
Niko
Attachments (1)
Change History (14)
comment:1 Changed 12 years ago by jmoore
- Cc m.woodbridge@… julien.jourde@… added
comment:2 Changed 12 years ago by jburel
- Sprint changed from 2013-02-12 (5) to 2013-03-12 (6))
Moved from sprint 2013-02-12 (5)
comment:3 Changed 12 years ago by jburel
- Sprint changed from 2013-03-12 (6)) to 2013-04-09 (7))
Moved from sprint 2013-03-12 (6))
comment:4 Changed 12 years ago by wmoore
Chris - having discussed this with Blazej and Kenny, we're no closer to understanding what the fundamental issue (and fix) is. Not sure if the problems above are caused by deployment settings or code bugs. The urls within the webstart/jars/insight.jnlp are generated with the same http / https prefix as the request.
Let me know if we need to fix the code, or improve docs or neither.
Cheers,
comment:5 Changed 11 years ago by jburel
- Sprint changed from 2013-04-09 (7)) to 2013-05-07 (8)
Moved from sprint 2013-04-09 (7))
comment:6 Changed 11 years ago by wmoore
- Owner changed from wmoore to cxallan
Chris, could you have a look at this (see comments above)? Cheers,
comment:7 Changed 11 years ago by jburel
- Sprint changed from 2013-05-07 (8) to Blocker 4.4.9 (1)
Moved from sprint 2013-05-07 (8)
comment:8 Changed 11 years ago by atarkowska
- Cc atarkowska added
- Version set to 4.4.8
comment:9 Changed 11 years ago by jamoore
- Owner changed from cxallan to atarkowska
Passing to Ola for evaluation.
comment:10 Changed 11 years ago by atarkowska
I've been playing with that issue, but I can't reproduce it.
We have SSL on on one of ours teaching environments and we do not experience that issue:
<?xml version="1.0" encoding="utf-8"?> <jnlp codebase="https://learning.openmicroscopy.org/dundee/static/webstart/jars/" href="https://learning.openmicroscopy.org/dundee/webstart/jars/insight.jnlp"> <information> <title>OMERO.insight</title> ... <j2se version="1.5+" max-heap-size="1024m"/> <jar href="omero.insight.jar"/> ... <jar href="jaxrpc.jar"/> <property name="jnlp.omero.host" value="localhost"/> <property name="jnlp.omero.port" value="4064"/> </resources> <application-desc main-class="org.openmicroscopy.shoola.Main"/> </jnlp>
and nginx config
server { listen 80 default; server_name ... rewrite ^ https://$server_name$request_uri? permanent; } server { listen 443 default ssl; ssl on; ssl_certificate ... ... location /omero/ { ... fastcgi_param HTTPS on;
Changed 11 years ago by atarkowska
comment:11 Changed 11 years ago by atarkowska
- Resolution set to worksforme
- Status changed from new to closed
Closing for now as it works for me
comment:12 Changed 11 years ago by atarkowska
It looks like fastcgi_param HTTPS was missed. See here
comment:13 Changed 10 years ago by khgillen
We should include a commented out HTTPS section in the nginx config which includes this, upon bin/omero web nginx.
Seehttp://lists.openmicroscopy.org.uk/pipermail/ome-users/2013-February/003515.html