requeuing chunks of the spool?

bill davidsen davidsen at tmr.com
Wed Jun 20 19:33:48 UTC 2001


In article <5.1.0.14.2.20010615103730.03ae0cf0 at lennier.cc.vt.edu>,
Ron Jarrell  <jarrell at vt.edu> wrote:

| Well, they finally listened to my comments that our old server was too slow, and had way
| to little disk space, and got me some new hardware, and I've gotten 2.3.2 built on it, with
| ovdb, and mostly cnfs.  (The old server was running cnfs as well).  I wanted to requeue
| about the last 180 days worth of our local groups, which are all in cnfs buffers, and
| send them over to the new system.  Last time I did this, I just used newsrequeue and 
| fed it the history file.  Which would work fine now, except that the 2.2 newsrequeue
| is still expecting message-ids, and the history file hasn't had plain text message ids for a
| while.  And the 2.3.2 newsrequeue solved that by only supporting log files (and I don't
| have 6 months of log files...)
| 
| So any suggestions on a straightforward way to requeue the last N days of the hierarchies
| I'm interested in?

  If this is a reader machine it's pretty easy. Get the XOVER for each
group (why I wrote getover), extract the message-id, run that through 
"grephistory -s" and out pops what you need for innxmit.

  getover oldsvr alt.sex.* *girl*soouts* *ford* | cut -f5 |
  grephistory -s >q.tmp
  innxmit -a -v newsvr $PWD/q.tmp

Ya, I do stuff like this, far too often.

-- 
bill davidsen <davidsen at tmr.com>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.


More information about the inn-workers mailing list