Timeout completing SSL handshake

[andreas.eriksson]

I'm trying to index our public web site using the Thunderstone Search Appliance, but I keep getting stuck on the very first page. The error reads:

Timeout completing SSL handshake with ssl.google-analytics.com:443

The problem is most likely that the search appliance is placed on the DMZ and the firewall is blocking the outbound https traffic.

How can I make SBE ignore any SSL references in the indexed pages?

[mark]

If they are "https" links uncheck "https" under allowed protocols on the all walk settings page.

If not you'll have to exclude them by url pattern. If the url has :443 in it you could use that as the exclusion. Otherwise list the undesired site names in the exclusions.

[andreas.eriksson]

No luck.

I disallowed HTTPS and added "google-analytics" to the Exclusions list. The only effect is that SBE seems to break all operation when it hits the google-analytics link, thereby ignoring all other links on the page.

Any suggestions?

-----------------------------------------
Walk started at 2008-01-29 13:17:37 (by resume)
JavaScript walking enabled
HTTPS walking disabled
Start fetching at http://www.stenaline.se/
Ignore urls containing any of the following:
/cgi-bin/
~
?
google-analytics
2008-01-29 13:17:37 started 1 new (12687) on http://www.stenaline.se/

015 /usr/local/morph3/texis/scripts/dowalk(doprimer) 401: Disallowed protocol `https' for URL https://ssl.google-analytics.com/urchin.js

015 /usr/local/morph3/texis/scripts/dowalk(doprimer) 401: Disallowed protocol `https' for URL https://ssl.google-analytics.com/urchin.js
0 pages fetched (94,912 bytes) from http://www.stenaline.se/
1 errors
0 duplicate pages
No pages fetched. Search not updated.
-------------------------------------------

[mark]

Set "Embedded Security" to "Same protocol".

[andreas.eriksson]

Works like a charm. Thank you very much, mark!