Hi Scott,
> The second started off OK then bogged down (dots slowed to a
> crawl). Eventually there was no noticable activity.
How much memory has your machine?
> [SNIP]
>
> bayes_toks: 2372553 keys copied
>
> That's it. No command prompt. It's *still* running!
I don't get it. There is no executable code between the line:
print STDERR "\n$file: $count keys copied\n";
and the lines:
print STDERR "done.\n";
exit;
So there is no reason why it should get stuck there. It must have to do
with external parameters, like harddisk activity (swapping!?), high CPU
load, etc. Did you notice anything of that sort?
What OS / Kernel are you running?
> 57734 p0 DL+ 3:24.92 /usr/bin/perl -w ./db-2-text2.pl -i bayes_toks
Where (and when) did this line come from???
Did you check whether the process was still running?
(something like: ps -alef | grep db-to-text2)
> > maybe try to delete the old bayes_toks before running the 2nd step.
>
> Tried that before but will try it again if you wish.
Last time you said that bayes_toks was still the old version afterwards.
When you first deleted it did the script recreate it? Was there any sign
at all of a new bayes_toks?
> >What are the last 10 lines of bayes_toks.txt?
>
> <key>gotham</key><ts>21</ts><th>0</th><atime>1083805428</atime>
> <key>remove.asp</key><ts>175</ts><th>0</th><atime>1083803806</atime>
> <key> NTOKENS</key><val>2372545</val>
> <key> LASTJOURNALSYNC</key><val>1083804822</val>
> <key> NHAM</key><val>39183</val>
> <key> OLDESTAGE</key><val>0</val>
> <key> NSPAM</key><val>79222</val>
> <key> DBVERSION</key><val>2</val>
> <key> NEWESTAGE</key><val>1083805428</val>
> <key> RUNNINGEXPIRE</key><val>1083804826</val>
> <key> LASTEXPIRE</key><val>1083767470</val>
Are you sure there are spaces after "<key>"?
They should be like this:
<key>^M^A^G^I^CLASTJOURNALSYNC</key><val>1083789372</val>
<key>^M^A^G^I^CNEWESTAGE</key><val>1083789211</val>
<key>^M^A^G^I^CDBVERSION</key><val>2</val>
<key>^M^A^G^I^CNHAM</key><val>5773</val>
<key>^M^A^G^I^CNTOKENS</key><val>149435</val>
<key>^M^A^G^I^CLASTATIMEDELTA</key><val>2764800</val>
<key>^M^A^G^I^CLASTEXPIREREDUCE</key><val>17410</val>
<key>^M^A^G^I^CLASTEXPIRE</key><val>1083602076</val>
<key>^M^A^G^I^COLDESTAGE</key><val>1080817327</val>
<key>^M^A^G^I^CNSPAM</key><val>8283</val>
> >Did you sync the journal before doing the text export as I suggested?
>
> No. From what I was able to discern from 'man sa-learn', this is done
> when '--rebuild' is fed to sa-learn. You cannot, AFAIK, specify a
> different working directory.
Just copy your local.cf, change the "bayes_path" and point sa-learn to the
new location with --configpath=path.
Andy.
--
o _ _ _
------- __o __o /\_ _ \\o (_)\__/o (_) -o)
----- _`\<,_ _`\<,_ _>(_) (_)/<_ \_| \ _|/' \/ /\\
---- (_)/ (_) (_)/ (_) (_) (_) (_) (_)' _\o_ _\_v
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Computers are useless. They can only give answers. (Pablo Picasso)
