Page 1 of 1

walking gw on existant database fails

Posted: Tue Jan 27, 1998 6:31 pm
by Thunderstone


when we add new URL to existing database, gw is walking through old refs =
like these:

Getting http://194.8.12.142/robots.txt...Not there...Ok.
Getting http://195.62.128.36/robots.txt...Not there...Ok.
...

and when any of previously walked servers are down, gw stops and hangs =
up without comments.

so it is inpossible to walk new URL until all previously walked web =
servers are operational.

is it our mistake or is it possible to remove referencies to walked =
URLs?

I tried to find were they reside, but without results :(

thank you,
edmunds zalite=20




walking gw on existant database fails

Posted: Tue Jan 27, 1998 6:56 pm
by Thunderstone



If the server is not there gw will continue after a TCP timeout.
To eliminate your issue you can delete the list of "known" Urls from
the options table. This will also disable the ability to use -rewalk.

gw -st "delete from options where Name='URL'"

Also make sure you don't have any Urls left over in todo from aborted
previous walks.

gw -st "delete from todo"