Including the text of the error would be useful. It's unclear as of this moment where the root cause of the issue lies. I've attempted to reproduce using the code you supplied, but instead saw either no error, or one time, "deadline exceeded" from urlfetch, which could be caused by a high load on the server. A dig query from a GCE machine shows they have records in place, and curl can successfully retrieve a page from
rutracker.org.
Is this still happening to you?
dig rutracker.org ANY
; <<>> DiG 9.9.5-3ubuntu0.6-Ubuntu <<>> rutracker.org ANY
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 57269
;; flags: qr rd ra; QUERY: 1, ANSWER: 10, AUTHORITY: 0, ADDITIONAL: 1
;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 512
;; QUESTION SECTION:
;rutracker.org. IN ANY
;; ANSWER SECTION:
rutracker.org. 489 IN SOA ns1.rutracker.org. hw.rutracker.org. 15122121 3600 900 120960 600
rutracker.org. 489 IN NS ns2.rutracker.org.
rutracker.org. 489 IN NS ns4.rutracker.org.
rutracker.org. 489 IN NS ns1.rutracker.org.
rutracker.org. 489 IN NS ns3.rutracker.org.
rutracker.org. 489 IN NS ns5.rutracker.org.
rutracker.org. 489 IN A 195.82.146.214
rutracker.org. 489 IN AAAA 2a02:4680:22::214
rutracker.org. 489 IN MX 5 mail.rutracker.org.
rutracker.org. 489 IN TXT "v=spf1 mx mx:mail.rutracker.org -all"
;; Query time: 1 msec
;; SERVER: 169.254.169.254#53(169.254.169.254)
;; WHEN: Mon Dec 21 18:42:03 UTC 2015
;; MSG SIZE rcvd: 285