On 2009-06-28 02:14:05 -0700, Ryan Niebur wrote:
> problems such as what?
I don't know. For some reason, it seems that the DNS request from
bind times out after 5 seconds, but then the second DNS sever from
/etc/resolv.conf does the resolving. However this doesn't explain
the full freeze.
> I can assure you there isn't code in apt-listbugs that purposely
> freezes.
But it can be in some module it uses or in ruby itself.
> unless you can help us with this, we can't fix the bug.
> sorry. perhaps you have your apt stuff configured to use a proxy?
I have the default configuration, so, AFAIK, no proxy.
>From the strace output (that I had attached to this bug report):
16094 11:09:31 socket(PF_INET, SOCK_STREAM, IPPROTO_TCP <unfinished ...>
16094 11:09:31 <... socket resumed> ) = 3
16094 11:09:31 fcntl(3, F_GETFL) = 0x2 (flags O_RDWR)
16094 11:09:31 fcntl(3, F_SETFL, O_RDWR|O_NONBLOCK) = 0
16094 11:09:31 connect(3, {sa_family=AF_INET, sin_port=htons(80),
sin_addr=inet_addr("140.211.166.43")}, 16) = -1 EINPROGRESS (Operation now in
progress)
16094 11:09:31 select(5, [], [3], [3], {993, 965542} <unfinished ...>
16094 11:09:32 <... select resumed> ) = 1 (out [3], left {993, 758585})
16094 11:09:32 connect(3, {sa_family=AF_INET, sin_port=htons(80),
sin_addr=inet_addr("140.211.166.43")}, 16) = 0
16094 11:09:32 fcntl(3, F_SETFL, O_RDWR) = 0
16094 11:09:32 fcntl(3, F_GETFL) = 0x2 (flags O_RDWR)
16094 11:09:32 fstat(3, {st_mode=S_IFSOCK|0777, st_size=0, ...}) = 0
16094 11:09:32 mmap(NULL, 4096, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f0b051ca000
16094 11:09:32 lseek(3, 0, SEEK_CUR) = -1 ESPIPE (Illegal seek)
16094 11:09:32 fcntl(3, F_GETFL) = 0x2 (flags O_RDWR)
16094 11:09:32 fstat(3, {st_mode=S_IFSOCK|0777, st_size=0, ...}) = 0
16094 11:09:32 mmap(NULL, 4096, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f0b051c7000
16094 11:09:32 lseek(3, 0, SEEK_CUR) = -1 ESPIPE (Illegal seek)
16094 11:09:32 select(0, [], [], [], {0, 0}) = 0 (Timeout)
16094 11:09:32 rt_sigprocmask(SIG_BLOCK, NULL, [], 8) = 0
16094 11:09:32 getsockname(3, {sa_family=AF_INET, sin_port=htons(33973),
sin_addr=inet_addr("152.81.65.165")}, [944242787589554192]) = 0
[...]
16094 11:09:37 select(0, [], [], [], {0, 0}) = 0 (Timeout)
16094 11:09:37 getsockname(3, {sa_family=AF_INET, sin_port=htons(33973),
sin_addr=inet_addr("152.81.65.165")}, [363347155135496208]) = 0
[...]
16094 11:09:42 rt_sigprocmask(SIG_BLOCK, NULL, [], 8) = 0
16094 11:09:42 select(5, [], [3], [], {998, 997769}) = 1 (out [3], left {998,
997760})
16094 11:09:42 select(5, [], [3], [], {998, 997641}) = 1 (out [3], left {998,
997635})
16094 11:09:42 write(3, "POST /cgi-bin/soap.cgi HTTP/1.1\r"..., 33) = 33
[...]
16094 11:09:42 select(5, [], [3], [], {998, 994828}) = 1 (out [3], left {998,
994822})
16094 11:09:42 select(5, [], [3], [], {998, 994706}) = 1 (out [3], left {998,
994700})
16094 11:09:42 write(3, "SOAPAction: \"\"\r\nContent-Type: te"..., 211) = 211
16094 11:09:42 select(5, [], [3], [], {998, 994497} <unfinished ...>
16094 11:09:42 <... select resumed> ) = 1 (out [3], left {998, 994491})
16094 11:09:42 select(0, [], [], [], {0, 0}) = 0 (Timeout)
16094 11:09:42 select(5, [], [3], [], {998, 994186}) = 1 (out [3], left {998,
994180})
16094 11:09:42 write(3, "\r\n", 2) = 2
16094 11:09:42 rt_sigprocmask(SIG_BLOCK, NULL, [], 8) = 0
16094 11:09:42 rt_sigprocmask(SIG_BLOCK, NULL, [], 8) = 0
16094 11:09:42 rt_sigprocmask(SIG_BLOCK, NULL, [], 8) = 0
16094 11:09:42 select(5, [], [3], [], {998, 993742}) = 1 (out [3], left {998,
993736})
16094 11:09:42 select(5, [], [3], [], {998, 993625}) = 1 (out [3], left {998,
993619})
16094 11:09:42 write(3, "<?xml version=\"1.0\" encoding=\"ut"..., 1053) = 1053
[...]
16094 11:09:42 select(4, [3], [], [], {998, 999347} <unfinished ...>
At this time, it freezes. The value used for the timeout is huge.
It should not be more than 2 minutes without some feedback.
Also there are a couple of "Illegal seek" on fd 3 above.
--
Vincent Lefèvre <vinc...@vinc17.org> - Web: <http://www.vinc17.org/>
100% accessible validated (X)HTML - Blog: <http://www.vinc17.org/blog/>
Work: CR INRIA - computer arithmetic / Arenaire project (LIP, ENS-Lyon)
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
