[Mailman-Users] FW: Output from "cron" command

[Mark Sapiro]

Jim Popovitch wrote:

>Mark Sapiro wrote:
>> 
>> I assume you are no longer getting errors from cron/checkdbs. What
>> happens if you put the request.pck back and try to go to the admindb
>> page for the list? Does this produce the error? 
>
>Yes.
>
>> If not, does it update the request.pck? If so, this may fix the problem.
>
>If I delete the request.pck files Mailman recreates them, but then the 
>admin pages show no pending issues.


I meant what if you restore the old request.pck and then go to the
admindb page for the list.


>> If none of this helps, or if you have a new request.pck with requests
>> since removing the old one, you could try bin/dumpdb on the
>> request.pck files to see what's in them, although I'm not sure how you
>> would use this information if you couldn't use the files themselves.
>
>bin/dumpdb shows me the data from the various request.pck files, what is 
>the format of entires, presumably they are subscription/post requests?


Yes, they are. The file is a pickled dictionary with keys and values.
The keys are generally numbers called msgids (not to be confused with
email Message-Id:) and the values are 2-tuples consisting of an
operation and data.

There is a special key = 'version' with data (0, 1), i.e. operation = 0
and data = version number currently = 1.

The rest of the keys are numeric and the operations are numbers as
follows:
1 -> held message
2 -> subscription request
3 -> unsub request

The data varies according to the operation.

For a held message, data is a 6-tuple consisting of time, sender,
subject, hold reason, message filename and message metadata dictionary.

For a sub, data is a 6-tuple consisting of time, address, full name,
password, digest flag and language.

For an unsub, data is just the address.

See Mailman/ListAdmin.py for more detail.

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