Open GoogleCodeExporter opened 8 years ago
also i have to add that i see lots of
2009/04/03 10:37:25| httpReadReply: FD 214: skipping 1xx response!
2009/04/03 10:37:25| httpReadReply: FD 214: skipping 1xx response!
2009/04/03 10:37:27| httpReadReply: FD 111: skipping 1xx response!
2009/04/03 10:37:27| httpReadReply: FD 413: skipping 1xx response!
2009/04/03 10:37:28| httpReadReply: FD 413: skipping 1xx response!
2009/04/03 10:37:28| httpReadReply: FD 111: skipping 1xx response!
2009/04/03 10:37:28| httpReadReply: FD 111: skipping 1xx response!
2009/04/03 10:37:30| httpReadReply: FD 111: skipping 1xx response!
2009/04/03 10:37:35| httpReadReply: FD 229: skipping 1xx response!
2009/04/03 10:37:35| httpReadReply: FD 267: skipping 1xx response!
2009/04/03 10:37:35| httpReadReply: FD 267: skipping 1xx response!
2009/04/03 10:37:35| httpReadReply: FD 229: skipping 1xx response!
2009/04/03 10:37:39| httpReadReply: FD 513: skipping 1xx response!
2009/04/03 10:37:39| httpReadReply: FD 389: skipping 1xx response!
2009/04/03 10:37:39| httpReadReply: FD 389: skipping 1xx response!
2009/04/03 10:37:39| httpReadReply: FD 513: skipping 1xx response!
2009/04/03 10:37:40| httpReadReply: FD 111: skipping 1xx response!
2009/04/03 10:37:42| httpReadReply: FD 386: skipping 1xx response!
2009/04/03 10:37:43| httpReadReply: FD 386: skipping 1xx response!
2009/04/03 10:37:44| httpReadReply: FD 322: skipping 1xx response!
2009/04/03 10:37:44| httpReadReply: FD 563: skipping 1xx response!
2009/04/03 10:37:44| httpReadReply: FD 322: skipping 1xx response!
2009/04/03 10:37:45| httpReadReply: FD 563: skipping 1xx response!
2009/04/03 10:37:46| httpReadReply: FD 244: skipping 1xx response!
2009/04/03 10:37:46| httpReadReply: FD 300: skipping 1xx response!
2009/04/03 10:37:47| httpReadReply: FD 244: skipping 1xx response!
why ? and what are these ????
besides i had to disable-htcp in order to get rid of the ctx-level 0 stuff
regards
Original comment by degreane@gmail.com
on 3 Apr 2009 at 7:45
Try applying this diff:
Index: http.c
===================================================================
--- http.c (revision 13904)
+++ http.c (working copy)
@@ -1125,7 +1125,7 @@
/* Did we get a successful parse but 1xx? Try again */
if (reply->sline.status >= 100 && reply->sline.status < 200) {
- debug(1, 1) ("httpReadReply: FD %d: skipping 1xx response!\n",
fd);
+ debug(1, 2) ("httpReadReply: FD %d: skipping 1xx response!\n",
fd);
httpReplyReset(reply);
httpState->reply_hdr_state = 0;
po += done; /* Skip the reply in the incoming buffer */
I wonder if the 1xx skipping code is properly working. I need to find a site
which is triggering that particular
error. Hm..
Original comment by adrian.c...@gmail.com
on 3 Apr 2009 at 7:57
ok but this doesn't actually fix the error ...mmmmmm
what about the high ratio of near hits ????
regards
Original comment by degreane@gmail.com
on 3 Apr 2009 at 8:15
and still the dns stat doesn't show ??? why ????
regards
Original comment by degreane@gmail.com
on 3 Apr 2009 at 8:23
re dns stats - http://code.google.com/p/lusca-cache/issues/detail?id=8
re - near hits? no idea. Will look at that next.
re - "error" - does it quieten the logging in cache.log ? Excessive cache.log
logging will affect throughput quite
significantly..
Original comment by adrian.c...@gmail.com
on 3 Apr 2009 at 11:32
Try this:
Index: http.c
===================================================================
--- http.c (revision 13911)
+++ http.c (working copy)
@@ -1125,7 +1125,10 @@
/* Did we get a successful parse but 1xx? Try again */
if (reply->sline.status >= 100 && reply->sline.status < 200) {
+ Ctx ctx = ctx_enter(entry->mem_obj->url);
debug(1, 1) ("httpReadReply: FD %d: skipping 1xx response!\n", fd);
+ debug(1, 1) ("httpReadReply: FD %d: reply
was:\n===\n%.*s\n====\n", fd, done,
buf_buf(httpState->read_buf) + po);
+ ctx_exit(ctx);
httpReplyReset(reply);
httpState->reply_hdr_state = 0;
po += done; /* Skip the reply in the incoming buffer */
I specifically want to see what URLs you're fetching from so I can try and
reproduce it locally!
Original comment by adrian.c...@gmail.com
on 5 Apr 2009 at 8:38
Original issue reported on code.google.com by
degreane@gmail.com
on 3 Apr 2009 at 7:26