Hi.
I've updated to mailman 2.1.3 on my rh7.3 system. The lists
work. Most of the admin functions work. Check_perms says everything is ok. However, the functions available from the admindb CGI don't work. They didn't in 2.1-8 either. There seems to be no error log I can locate. I know admindb is running because I moved it and the Web server complained, so it is looking at the right CGI. But it doesn't do any of the four available status changes when I try them on any message in any list. This would seem to indicate a systemic problem in the CGI. The group ownership is correct as is the SGID setting.
Are there any diagnostics I can turn on in mailman? Does anyone
have ANY clue what might be going on here? I have a backlog of held messages I'd very much like to clear up.
Help GREATLY appreciated.
Chris Johnson |Internet: johnson@nmr.mgh.harvard.edu Systems Administrator |Web: http://www.nmr.mgh.harvard.edu/~johnson NMR Center |Voice: 617.726.0949 Mass. General Hospital |FAX: 617.726.7422 149 (2301) 13th Street |"Life is chaos. Chaos is life. Control is an Charlestown, MA., 02129 USA | illusion." Trance Gemini
On Mon, 2003-12-22 at 09:25, Chris Johnson wrote:
Hi. I've updated to mailman 2.1.3 on my rh7.3 system. The lists
work. Most of the admin functions work. Check_perms says everything is ok. However, the functions available from the admindb CGI don't work. They didn't in 2.1-8 either. There seems to be no error log I can locate. I know admindb is running because I moved it and the Web server complained, so it is looking at the right CGI. But it doesn't do any of the four available status changes when I try them on any message in any list. This would seem to indicate a systemic problem in the CGI. The group ownership is correct as is the SGID setting.
Are there any diagnostics I can turn on in mailman? Does anyone
have ANY clue what might be going on here? I have a backlog of held messages I'd very much like to clear up.
Help GREATLY appreciated.
I've seen this a few times, most often when the webserver using a redirect for the URL. Also, make sure that you have the python2-dev rpm installed.
Good Luck - Jon Carnes
On 22 Dec 2003, Jon Carnes wrote:
On Mon, 2003-12-22 at 09:25, Chris Johnson wrote:
Hi. I've updated to mailman 2.1.3 on my rh7.3 system. The lists
work. Most of the admin functions work. Check_perms says everything is ok. However, the functions available from the admindb CGI don't work. They didn't in 2.1-8 either. There seems to be no error log I can locate. I know admindb is running because I moved it and the Web server complained, so it is looking at the right CGI. But it doesn't do any of the four available status changes when I try them on any message in any list. This would seem to indicate a systemic problem in the CGI. The group ownership is correct as is the SGID setting.
Are there any diagnostics I can turn on in mailman? Does anyone
have ANY clue what might be going on here? I have a backlog of held messages I'd very much like to clear up.
Help GREATLY appreciated.
I've seen this a few times, most often when the webserver using a redirect for the URL. Also, make sure that you have the python2-dev rpm installed.
Good Luck - Jon Carnes
I have
python2-2.2.2-11.7.3 python2-devel-2.2.2-11.7.3
in the system. And I'm running this directly off the system the lsts are on. So there are no URL redirects.
If the python2 I'm using is any good then at least I think I know
what it isn't. Hopefully.
Thank you. Any other ideas?
I'm desperate enough that I'm tracking code. If you're familiar
with it at all, the process_form fuction never gets called. If I'm reading the code right, admindb thinks my form submission isn't a form submission. Very strange. I'm running the apache-1.3.27-3 rpm also.
Chris Johnson |Internet: johnson@nmr.mgh.harvard.edu Systems Administrator |Web: http://www.nmr.mgh.harvard.edu/~johnson NMR Center |Voice: 617.726.0949 Mass. General Hospital |FAX: 617.726.7422 149 (2301) 13th Street |Doctors don't save lives. The best they can hope Charlestown, MA., 02129 USA |to do is save life. Not the same thing. Me
On Mon, 2003-12-22 at 10:59, Chris Johnson wrote:
On 22 Dec 2003, Jon Carnes wrote:
On Mon, 2003-12-22 at 09:25, Chris Johnson wrote:
Hi. I've updated to mailman 2.1.3 on my rh7.3 system. The lists
work. Most of the admin functions work. Check_perms says everything is ok. However, the functions available from the admindb CGI don't work. They didn't in 2.1-8 either. There seems to be no error log I can locate. I know admindb is running because I moved it and the Web server complained, so it is looking at the right CGI. But it doesn't do any of the four available status changes when I try them on any message in any list. This would seem to indicate a systemic problem in the CGI. The group ownership is correct as is the SGID setting.
Are there any diagnostics I can turn on in mailman? Does anyone
have ANY clue what might be going on here? I have a backlog of held messages I'd very much like to clear up.
Help GREATLY appreciated.
I've seen this a few times, most often when the webserver using a redirect for the URL. Also, make sure that you have the python2-dev rpm installed.
Good Luck - Jon Carnes
I have
python2-2.2.2-11.7.3 python2-devel-2.2.2-11.7.3
in the system. And I'm running this directly off the system the lsts are on. So there are no URL redirects.
If the python2 I'm using is any good then at least I think I know
what it isn't. Hopefully.
Thank you. Any other ideas? I'm desperate enough that I'm tracking code. If you're familiar
with it at all, the process_form fuction never gets called. If I'm reading the code right, admindb thinks my form submission isn't a form submission. Very strange. I'm running the apache-1.3.27-3 rpm also.
Those rpms should give you all the Python you need.
Have you tried "VIRTUAL_HOST_OVERVIEW = 0" in your mm_cfg.py file? What does your ~mailman/Mailman/mm_cfg.py file look like? Can you include a copy.
Also, have your run "check_perms" on the system yet?
Jon Carnes
On 22 Dec 2003, Jon Carnes wrote:
Those rpms should give you all the Python you need.
Have you tried "VIRTUAL_HOST_OVERVIEW = 0" in your mm_cfg.py file?
Uh, nope..... Um, maybe I'm doing something wrong but I don't
see any difference in listings. What should I be looking for?
What does your ~mailman/Mailman/mm_cfg.py file look like? Can you include a copy.
The relevent adds follow.
from Defaults import * import pwd, grp
############################################################## # Here's where we override shipped defaults with settings # # suitable for the RPM package. # MAILMAN_UID = pwd.getpwnam('mailman')[2] MAILMAN_GID = grp.getgrnam('mailman')[2] MAILMAN_USER = 'mailman' MAILMAN_GROUP = 'mailman' LOG_DIR = '/var/log/mailman' #QUEUE_DIR = '/var/spool/mailman/qfiles'
############################################################## # Put YOUR site-specific configuration below, in mm_cfg.py . # # See Defaults.py for explanations of the values. #
DEFAULT_EMAIL_HOST = 'nmr.mgh.harvard.edu' DEFAULT_URL_HOST = 'www.nmr.mgh.harvard.edu' MAILMAN_OWNER = 'mailman-owner@%s' % DEFAULT_HOST_NAME
# Note - if you're looking for something that is imported from mm_cfg, but you # didn't find it above, it's probably in Defaults.py. # # add_virtualhost(DEFAULT_URL_HOST, DEFAULT_EMAIL_HOST) # # #add_virtualhost(DEFAULT_URL_HOST, DEFAULT_EMAIL_HOST) # DEFAULT_LIST_ADVERTISED = 0 # DEFAULT_ADMIN_NOTIFY_MCHANGES = 1 # DEFAULT_GENERIC_NONMEMBER_ACTION = 2
Also, have your run "check_perms" on the system yet?
Yep. Looks clean.
Jon Carnes
Chris Johnson |Internet: johnson@nmr.mgh.harvard.edu Systems Administrator |Web: http://www.nmr.mgh.harvard.edu/~johnson NMR Center |Voice: 617.726.0949 Mass. General Hospital |FAX: 617.726.7422 149 (2301) 13th Street |I've heard men and women say all the good ones are Charlestown, MA., 02129 USA |taken. What does that make the rest of us? Me
Jon,
Are you familiar with the code at all in admindb? I'm tracing
back up from line 150. Again, I don't really know python, but if I'm reading this right, the line where it's supposed to grab the QUERY_STRING from the HTTP environment is completely empty.
Wierd!
Chris Johnson |Internet: johnson@nmr.mgh.harvard.edu Systems Administrator |Web: http://www.nmr.mgh.harvard.edu/~johnson NMR Center |Voice: 617.726.0949 Mass. General Hospital |FAX: 617.726.7422 149 (2301) 13th Street |For all sad words of tongue or pen, the saddest Charlestown, MA., 02129 USA |are these: "It might have been". John G. Whittier
On Mon, 2003-12-22 at 12:25, Chris Johnson wrote:
Jon, Are you familiar with the code at all in admindb? I'm tracing
back up from line 150. Again, I don't really know python, but if I'm reading this right, the line where it's supposed to grab the QUERY_STRING from the HTTP environment is completely empty.
Wierd!
I'm not an expert, just a plain ordinary Python hacker, but it looks like all your data is being dumped (like in a redirect when all the posted data is dumped).
What do your Apache logs show when you do a transaction?
22-Dec-03 at 12:44, Jon Carnes (jonc@nc.rr.com) wrote :
On Mon, 2003-12-22 at 12:25, Chris Johnson wrote:
Jon, Are you familiar with the code at all in admindb? I'm tracing
back up from line 150. Again, I don't really know python, but if I'm reading this right, the line where it's supposed to grab the QUERY_STRING from the HTTP environment is completely empty.
Wierd!
I'm not an expert, just a plain ordinary Python hacker, but it looks like all your data is being dumped (like in a redirect when all the posted data is dumped).
What do your Apache logs show when you do a transaction?
Stab in the dark without logs, but I reckon it might be the sort of automatic URL rewriting in Apache that is doing this. It will often change the received hostname to the ServerName directive, esp. if there are no virtual hosts in the config and there are several hostnames that point to the box.
This usually ends up just redirecting to login and then all is OK, but maybe not?
-- Simon White. Internet Consultant, Linux/Windows Server Administration. email, dns and web servers; php javascript perl asp; MySQL MSSQL Access Bridging the gap between management, HR and the tech team.
On 22 Dec 2003, Jon Carnes wrote:
On Mon, 2003-12-22 at 12:25, Chris Johnson wrote:
Jon, Are you familiar with the code at all in admindb? I'm tracing
back up from line 150. Again, I don't really know python, but if I'm reading this right, the line where it's supposed to grab the QUERY_STRING from the HTTP environment is completely empty.
Wierd!
I'm not an expert, just a plain ordinary Python hacker, but it looks like all your data is being dumped (like in a redirect when all the posted data is dumped).
What do your Apache logs show when you do a transaction?
Hm, .... odd. Do you know where admindb gets the information to
produce the FORM ACTION= line?
Chris Johnson |Internet: johnson@nmr.mgh.harvard.edu Systems Administrator |Web: http://www.nmr.mgh.harvard.edu/~johnson NMR Center |Voice: 617.726.0949 Mass. General Hospital |FAX: 617.726.7422 149 (2301) 13th Street |Life n, something from which nobody escapes Charlestown, MA., 02129 USA |alive. Me
:: I've updated to mailman 2.1.3 on my rh7.3 system. The lists :: work. Most of the admin functions work. Check_perms says everything :: is ok. However, the functions available from the admindb CGI don't :: work.
Chris, if you do a dumpdb on a list that doesn't work, does the value of the 'web_page_url' option match the URL that you're using to access the admin page? You haven't said whether you're using SSL or not, but I had this very problem and it was due to the fact that I was using SSL and all the list URLs were referencing http:// instead of https://
Let me know if this applies to you.
Cheers - Erick
On Tue, 23 Dec 2003, Erick Mechler wrote:
:: I've updated to mailman 2.1.3 on my rh7.3 system. The lists :: work. Most of the admin functions work. Check_perms says everything :: is ok. However, the functions available from the admindb CGI don't :: work.
Chris, if you do a dumpdb on a list that doesn't work, does the value of the 'web_page_url' option match the URL that you're using to access the admin page? You haven't said whether you're using SSL or not, but I had this very problem and it was due to the fact that I was using SSL and all the list URLs were referencing http:// instead of https://
Let me know if this applies to you.
Cheers - Erick
Not SSL but something similar with lists moved to a different
system. Thanks.
Chris Johnson |Internet: johnson@nmr.mgh.harvard.edu Systems Administrator |Web: http://www.nmr.mgh.harvard.edu/~johnson NMR Center |Voice: 617.726.0949 Mass. General Hospital |FAX: 617.726.7422 149 (2301) 13th Street |"Survival is insufficient" Charlestown, MA., 02129 USA | Seven of nine.
:: Not SSL but something similar with lists moved to a different :: system. Thanks.
First you have to make sure that your settings for the following 3 variables are configured properly in your mm_cfg.py file:
DEFAULT_EMAIL_HOST = 'host.foo.net' DEFAULT_URL_HOST = 'host.foo.net' DEFAULT_URL_PATTERN = 'http://%s/mailman/'
(note that if you're using SSL, that should be https:// not http://)
Once that's done, run the following command to fix all the built-in URLs for all your lists:
~mailman/bin/withlist -l -r fix_url [listname] -v
You can use 'all' for [listname] to change this value for all your lists,
but I'd run it on one just to make sure you're doing the right thing.
After you run this, you can verify the change was made by running
~mailman/bin/dumpdb ~mailman/lists/[listname]/config.pck | grep http
After all that, the admin pages should work again.
Cheers - Erick
participants (4)
-
Chris Johnson
-
Erick Mechler
-
Jon Carnes
-
Simon White