On Wed, Aug 16, 2017 at 6:12 AM, sebb <seb...@gmail.com> wrote: > I think this is caused by the proxy server which sometimes returns an > error for a valid URL.
That's not how I read this stack trace. Mind you, the error makes no sense the way I do read it. The message says: bad URI(is not URI?): http://spamassassin.apache.org/ The cause for this message is that the uri passed to URI::RFC3986_Parser.split doesn't respond to "to_str". Which is a round-about way of saying that the uri isn't a String (or more precisely, doesn't behave like a String in a Duck typing sense). That uri is obtained by ASF::Site, and the value is obtained by parsing contents in svn working directories (i.e., no network access required - updates are done by cronjobs). And in nearly every run, is always a string. What's particularly puzzling is that I don't even see recursion here, which would indicate following a redirect or the like. What the message is clearly saying is that the uri isn't a string. And oddly enough, when the uri is included in the error message (under the covers using to_s instead of to_str), it looks right. > I'm not quite sure why a network access should should be triggered by > URI.parse(). > Whilst the code could retry (as for timeout) how can we then > distinguish bad URIs from a temporary proxy fail? > Ideally we don't want to retry repeatedly for invalid URIs until the > depth is exceeded, but I don't see how to distinguish them from the > proxy errors... > > On 16 August 2017 at 08:03, Ping My Box <no-re...@pingmybox.com> wrote: >> >> Hi there! >> The service at whimsy.apache.org (whimsy.apache.org (https)) appears to be >> down from multiple locations around the world. >> >> The exact error is: >> >> Error component: response >> Error code: Internal Server Error or equivalent bad message received: >> HTTP/1.1 400 site_scan >> ["/usr/local/rvm/rubies/ruby-2.4.1/lib/ruby/2.4.0/uri/rfc3986_parser.rb:18:in >> `rescue in split': bad URI(is not URI?): http://spamassassin.apache.org/ >> (URI::InvalidURIError)", "\tfrom >> /usr/local/rvm/rubies/ruby-2.4.1/lib/ruby/2.4.0/uri/rfc3986_parser.rb:15:in >> `split'", "\tfrom >> /usr/local/rvm/rubies/ruby-2.4.1/lib/ruby/2.4.0/uri/rfc3986_parser.rb:73:in >> `parse'", "\tfrom >> /usr/local/rvm/rubies/ruby-2.4.1/lib/ruby/2.4.0/uri/common.rb:231:in >> `parse'", "\tfrom /x1/srv/whimsy/lib/whimsy/cache.rb:104:in `fetch'", >> "\tfrom /x1/srv/whimsy/lib/whimsy/cache.rb:124:in `rescue in fetch'", >> "\tfrom /x1/srv/whimsy/lib/whimsy/cache.rb:101:in `fetch'", "\tfrom >> /x1/srv/whimsy/lib/whimsy/cache.rb:66:in `get'", "\tfrom site-scan.rb:28:in >> `parse'", "\tfrom site-scan.rb:184:in `block in <main>'", "\tfrom >> site-scan.rb:176:in `each'", "\tfrom site-scan.rb:176:in `<main>'"] >> Debug output: >> [Wed Aug 16 07:03:40 2017]: Initialising socket >> [Wed Aug 16 07:03:40 2017]: Looking up hostname whimsy.apache.org... >> [Wed Aug 16 07:03:40 2017]: Connecting to 54.91.229.41:443 >> [Wed Aug 16 07:03:41 2017]: Connected, sending HTTPS payload. >> [Wed Aug 16 07:03:41 2017]: Analyzing server certificate >> [Wed Aug 16 07:03:41 2017]: Saving certificate data >> [Wed Aug 16 07:03:41 2017]: Reading response header from server >> [Wed Aug 16 07:03:42 2017]: Caught exception: Internal Server Error or >> equivalent bad message received: HTTP/1.1 400 site_scan >> ["/usr/local/rvm/rubies/ruby-2.4.1/lib/ruby/2.4.0/uri/rfc3986_parser.rb:18:in >> `rescue in split': bad URI(is not URI?): http://spamassassin.apache.org/ >> (URI::InvalidURIError)", "\tfrom >> /usr/local/rvm/rubies/ruby-2.4.1/lib/ruby/2.4.0/uri/rfc3986_parser.rb:15:in >> `split'", "\tfrom >> /usr/local/rvm/rubies/ruby-2.4.1/lib/ruby/2.4.0/uri/rfc3986_parser.rb:73:in >> `parse'", "\tfrom >> /usr/local/rvm/rubies/ruby-2.4.1/lib/ruby/2.4.0/uri/common.rb:231:in >> `parse'", "\tfrom /x1/srv/whimsy/lib/whimsy/cache.rb:104:in `fetch'", >> "\tfrom /x1/srv/whimsy/lib/whimsy/cache.rb:124:in `rescue in fetch'", >> "\tfrom /x1/srv/whimsy/lib/whimsy/cache.rb:101:in `fetch'", "\tfrom >> /x1/srv/whimsy/lib/whimsy/cache.rb:66:in `get'", "\tfrom site-scan.rb:28:in >> `parse'", "\tfrom site-scan.rb:184:in `block in <main>'", "\tfrom >> site-scan.rb:176:in `each'", "\tfrom site-scan.rb:176:in `<main>'"] >> >> >> >> >> >> With regards, >> Ping My Box - https://www.pingmybox.com/
