The following error was always presented after index ran. It can be 
recreated every time running:

./index -D


OUTPUT:

Loading configuration from /usr/local/aspseek/etc/db.conf
Loading configuration from /usr/local/aspseek/etc/ucharset.conf
Loading configuration from /usr/local/aspseek/etc/stopwords.conf
Loading configuration from /usr/local/aspseek/etc/aspseek.conf
Saving real-time ... done
Saving redirects ... done
Saving direct href deleta files ... done
Calculating ranks [........] done.
Saving lastmods ... done
Generating word site ..Error: MySQL server has gone away <INSERT INTO 
wordsite(word, sites) VALUES(:1, :2)>
Error: MySQL server has gone away <INSERT INTO wordsite(word, sites) 
VALUES(:1, :2)>
. done

I don't know how long this has been going on because while running index 
this is usually scrolled past the screen when indexing is completed. This 
time I scrolled back to see the process and that's what I found after 
running index. To duplicate the same error I ran index -D and sure enough 
the error occured.

I'm running MySQL's my-huge.cnf with no modifications other than commenting 
out the log-bin. MySQL server has NOT gone away and it is live. No errors in 
MySQL's sitename.err file either and search works fine. No other users on 
this box just indexer and me. The box is not open to the world.

I have read information at MySQL's Website concerning this type of error. 
This page can be located at:

http://www.mysql.com/doc/en/Gone_away.html

I made a change to the my.cnf:

from:
set-variable = max_allowed_packet = 1m

to:

set-variable = max_allowed_packet = 3m

and that didn't help. So....

set-variable = max_allowed_packet = 10m

and it then was able to generate word site without an error.

My next step now is of course to find out what I need to set this at so I 
don't run into the same problem in the future. The maximum this can be set 
at using MySQL 3.23 is 16m. So now I'm wondering just how many URLs can we 
index if the packets are this large on just 2 million URLs? I also found 
this page to be interesting regarding this problem.

http://www.mysql.com/doc/en/Packet_too_large.html

Here are my current stats for index (I wish I could get rid of everything 
other than status 200 and status 0 though):


ASPseek database statistics

    Status    Expired      Total
   -----------------------------
         0      61298      61324 Not indexed yet
         1          0        132 Unknown status
       200          0    1974309 OK
       202          0         75 Unknown status
       204          0         45 No content
       205          0          3 Unknown status
       300          0         19 Multiple Choices
       301          0      62552 Moved Permanently
       302          0     121177 Moved Temporarily
       303          0          7 See Other
       307          0          8 Unknown status
       400          0        107 Bad Request
       401          0        181 Unauthorized
       402          0          5 Payment Required
       403          0       3115 Forbidden
       404          0    1129545 Not found
       405          0          1 Method Not Allowed
       407          0          1 Proxy Authentication Required
       408          0          5 Request Timeout
       410          0         10 Gone
       415          0          1 Unsupported Media Type
       500          0        344 Internal Server Error
       501          0         11 Not Implemented
       502          0         13 Bad Gateway
       503          0        328 Service Unavailable
       504          0          8 Gateway Timeout
       508          0        275 Unknown status
   -----------------------------
     Total      61298    3353601

Maybe Kir has an answer to this situation?

Thanks,
John


_________________________________________________________________
Chat with friends online, try MSN Messenger: http://messenger.msn.com

Reply via email to