Henrik Nordstrom wrote:
monitorurl needs to be a full URL, not just the url-path. I.e.
http://www.your.domain/sesame
OK, thank you, that clarifies the situation (though for originservers it
seems to work both ways, i.e. as directory or with full url).
This leaves me with the following pair of configurations:
cache_peer libby.ldeo.columbia.edu parent 8180 3130 no-query no-digest
originserver monitorinterval=30
monitorurl=http://libby.ldeo.columbia.edu:8180/sesame/ name=sesame_exp
round-robin
which works (though it is not quite what I need), and
cache_peer libby.ldeo.columbia.edu parent 8180 3130 no-query no-digest
originserver monitorinterval=30 monitorsize=100-200
monitorurl=http://libby.ldeo.columbia.edu:8180/sesame/servlets/evaluateGraphQuery?repository=dl-rdfs&serialization=ntriples&queryLanguage=SeRQL&query=CONSTRUCT+*%0D%0AFROM%0D%0A%7Br%3ARDFCache%7D+r%3Aupload_complete+%7Bflag%7D%0D%0AUSING+NAMESPACE%0D%0A+r+%3D+%3Chttp%3A%2F%2Firidl.ldeo.columbia.edu%2Fontologies%2Frdfcache.owl%23%3E
name=sesame_exp round-robin
which results in the server being marked down, and the following errors
in the cache.log file.
2007/01/04 11:26:56| peerMonitor sesame_exp: Failed, status != 200 (400)
2007/01/04 11:26:56| peerMonitor sesame_exp: Failed, status != 200 (400)
2007/01/04 11:27:26| peerMonitor sesame_exp: Failed, status != 200 (400)
2007/01/04 11:27:26| peerMonitor sesame_exp: Failed, status != 200 (400)
So it fails with the longer url, with a 400 error, which means the
client (squid) made a mistake.
Note too that the requests are in pairs, which is true even when it is
working (one can watch the fetch count in the cache-report go up by two
every 30 seconds). I can't tell if it is actually hitting the server
twice, however.
the url works entered in a browser. What am I missing?
Benno Blumenthal
--
Dr. M. Benno Blumenthal benno@xxxxxxxxxxxxxxxx
International Research Institute for climate and society
The Earth Institute at Columbia University
Lamont Campus, Palisades NY 10964-8000 (845) 680-4450
