Dennis Putnam wrote:

...

If you go to the admindb interface for the 'mailman' list, you should see these messages and be able to discard them - go to the same URL as the admindb interface for your other list, but with 'mailman' as the list name.

As I said, it tells me there are no pending requests.

This is a bit of a puzzle, because it should be using the same request.pck data as the cron/checkdbs job that sends the notice.

...

Or, you can discard them by running.

path/to/mailmans/bin/discard /var/lib/mailman/data/heldmsg-mailman-*

I thought I'd give this a try and here's what I get:

[root@dap002 bin]# discard /var/lib/mailman/data/heldmsg-mailman-* Traceback (most recent call last): File "/usr/sbin/discard", line 120, in ? main() File "/usr/sbin/discard", line 110, in main mlist.HandleRequest(id, mm_cfg.DISCARD, '', False, False, '') File "/usr/lib/mailman/Mailman/ListAdmin.py", line 164, in HandleRequest rtype, data = self.__db[id] KeyError: 113 [root@dap002 bin]#

Must be a remnant of the restore fiasco.

This says the heldmsg file did not have a corresponding request.pck entry, and presumably there are some the other way around too. I.e. after the restore, the actual heldmsg-* files were out of sync with request.pck entries.

...

Do NOT just remove the heldmsg-mailman-* files since that will remove the message files, but it won't stop the notices. If you have already removed the files, you need to either go to the admindb interface for the mailman list and discard them or remove the /var/lib/mailman/lists/mailman/request.pck file which will be automatically recreated when necessary.

Even though the above didn't work, can I do this anyway? I don't use that mailing list so there are no messages in it that I care about. Is there a way to just wipe everything manually? I guess this list is necessary but is there a way to wipe everything and have it rebuild what it needs? That should clean everything up, right?

I suggest you remove all the heldmsg-mailman-* files and also remove the lists/mailman/request.pck file. If you then go to the admindb interface for the 'mailman' list, it should create a new, 'empty' request.pck.

-- Mark Sapiro <mark@msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

Dennis Putnam wrote:

...

If you go to the admindb interface for the 'mailman' list, you should see these messages and be able to discard them - go to the same URL as the admindb interface for your other list, but with 'mailman' as the list name.

As I said, it tells me there are no pending requests.

This is a bit of a puzzle, because it should be using the same request.pck data as the cron/checkdbs job that sends the notice.

...

Or, you can discard them by running.

path/to/mailmans/bin/discard /var/lib/mailman/data/heldmsg-mailman-*

I thought I'd give this a try and here's what I get:

[root@dap002 bin]# discard /var/lib/mailman/data/heldmsg-mailman-* Traceback (most recent call last): File "/usr/sbin/discard", line 120, in ? main() File "/usr/sbin/discard", line 110, in main mlist.HandleRequest(id, mm_cfg.DISCARD, '', False, False, '') File "/usr/lib/mailman/Mailman/ListAdmin.py", line 164, in HandleRequest rtype, data = self.__db[id] KeyError: 113 [root@dap002 bin]#

Must be a remnant of the restore fiasco.

This says the heldmsg file did not have a corresponding request.pck entry, and presumably there are some the other way around too. I.e. after the restore, the actual heldmsg-* files were out of sync with request.pck entries.

...

Do NOT just remove the heldmsg-mailman-* files since that will remove the message files, but it won't stop the notices. If you have already removed the files, you need to either go to the admindb interface for the mailman list and discard them or remove the /var/lib/mailman/lists/mailman/request.pck file which will be automatically recreated when necessary.

Even though the above didn't work, can I do this anyway? I don't use that mailing list so there are no messages in it that I care about. Is there a way to just wipe everything manually? I guess this list is necessary but is there a way to wipe everything and have it rebuild what it needs? That should clean everything up, right?

I suggest you remove all the heldmsg-mailman-* files and also remove the lists/mailman/request.pck file. If you then go to the admindb interface for the 'mailman' list, it should create a new, 'empty' request.pck.

-- Mark Sapiro <mark@msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan