Open GoogleCodeExporter opened 8 years ago
If you disable an ICP peer.
cache.log
2009/06/22 22:18:10| disabling (peer disappeared) digest from A.B.C.D
BT
#0 0x080b7008 in storeLog (tag=3, e=0x76db7680) at store_log.c:68
#1 0x080af8c8 in storeRelease (e=0x76db7680) at store.c:1627
#2 0x080afda3 in storeUnlockObjectDebug (e=0x76db7680, file=0x8109969
"peer_digest.c", line=702) at store.c:333
#3 0x0809b645 in peerDigestReqFinish (fetch=0x77225bf8, fcb_valid=1,
pdcb_valid=1,
pcb_valid=0, reason=0x8109b62 "peer disappeared", err=1) at peer_digest.c:702
#4 0x0809bb59 in peerDigestFetchedEnough (fetch=0x77225bf8, size=1281,
step_name=0x8109bd0 "peerDigestFetchReply") at peer_digest.c:585
#5 0x0809beab in peerDigestFetchReply (data=0x77225bf8, nr={node = 0x7a44b7d0,
offset = 0}, size=1281) at peer_digest.c:355
#6 0x080b2e2c in storeClientCallback (sc=0x77225fb0, sz=-1) at
store_client.c:267
#7 0x080b3224 in storeClientCopy2 (e=0x77225e08, sc=0x77225fb0) at
store_client.c:520
#8 0x080b3512 in InvokeHandlers (e=0x77225e08) at store_client.c:870
#9 0x080adfa7 in storeBufferFlush (e=0x77225e08) at store.c:2048
#10 0x08076820 in errorAppendEntry (entry=0x77225e08, err=0x74203568) at
errorpage.c:331
#11 0x080b06c2 in storeRequestFailed (e=0x77225e08, err=0x74203568) at
store.c:1468
#12 0x0807a267 in fwdStateFree (fwdState=0x77226010) at forward.c:104
#13 0x080e409e in commCallCloseHandlers (fd=570) at comm.c:632
#14 0x080e4d13 in comm_close (fd=570) at comm.c:809
#15 0x080729db in commConnectCallback (cs=0x7a39bf40, status=-7) at comm.c:142
#16 0x08072aad in commConnectHandle (fd=570, data=0x7a39bf40) at comm.c:283
#17 0x080e9151 in comm_select (msec=10) at comm_generic.c:300
#18 0x080942e4 in main (argc=2, argv=0xbfeb0714) at main.c:934
Original comment by Fraw...@gmail.com
on 22 Jun 2009 at 12:31
This is almost certainly going to be method_t related bugs from Squid-2.HEAD.
I'm
going to have to add in some extra debugging to catch NULL method pointers
around the
code.
Thanks.
Original comment by adrian.c...@gmail.com
on 1 Jul 2009 at 11:29
Original comment by adrian.c...@gmail.com
on 1 Jul 2009 at 11:29
.. and the second backtrace is due to cache digests, which I just haven't really
tested in any meaningful way.
I'll first look at the ICP specific crash.
Original comment by adrian.c...@gmail.com
on 27 Jul 2009 at 3:19
.. this should be resolved somewhat in LUSCA_HEAD. Please test and tell me
whether you have the issue now.
Original comment by adrian.c...@gmail.com
on 27 Oct 2009 at 8:21
Original issue reported on code.google.com by
Fraw...@gmail.com
on 22 Jun 2009 at 12:26