[
https://issues.apache.org/jira/browse/TS-3311?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14345654#comment-14345654
]
Leif Hedstrom commented on TS-3311:
-----------------------------------
The patch looks fine. I haven't tried to reproduce the bug with it, but my
observations on this triggering is as follow:
1) The recursive resolve needs to resolve "localhost".
2) parent.config has to exist. I'm not 100% certain, but I don't think it has
to be enabled (we parse parent.config regardless if it's enabled or not).
3) It'll only trigger once the entry for "localhost" is in the HostDB cache,
and the TTL expires. I.e. on revalidation of the entry.
4) It seems to primarily happen on the traffic_cop health checks (i.e. proxying
http://127.0.0.1:8083/synthetic.txt).
> Possible lookups on NULL hostnames in HostDB
> --------------------------------------------
>
> Key: TS-3311
> URL: https://issues.apache.org/jira/browse/TS-3311
> Project: Traffic Server
> Issue Type: Bug
> Components: HostDB
> Reporter: Steve Malenfant
> Assignee: Alan M. Carroll
> Fix For: 5.3.0
>
>
> Getting multiple segfaults per day on 4.2.1.
> [4324544.324222] [ET_NET 23][10504]: segfault at 0 ip 00002acd66546168 sp
> 00002acd71f190b8 error 4 in libtsutil.so.4.2.1[2acd66521000+34000]
> [4410696.820857] [ET_NET 19][22738]: segfault at 0 ip 00002af09f339168 sp
> 00002af0aa9230b8 error 4 in libtsutil.so.4.2.1[2af09f314000+34000]
> [4497039.474253] [ET_NET 12][34872]: segfault at 0 ip 00002ad17e6a1168 sp
> 00002ad1896100b8 error 4 in libtsutil.so.4.2.1[2ad17e67c000+34000]
> [4583372.073916] [ET_NET 3][46994]: segfault at 0 ip 00002aced4227168 sp
> 00002aceda7d80b8 error 4 in libtsutil.so.4.2.1[2aced4202000+34000]
> [4756046.944373] [ET_NET 22][10799]: segfault at 0 ip 00002b1771f76168 sp
> 00002b177d9130b8 error 4 in libtsutil.so.4.2.1[2b1771f51000+34000]
> Stack Trace :
> (gdb) bt
> #0 ink_inet_addr (s=<value optimized out>) at ink_inet.cc:107
> #1 0x00000000005e0df5 in is_dotted_form_hostname (mutex=0x1d32cb0, md5=...,
> ignore_timeout=false) at P_HostDBProcessor.h:545
> #2 probe (mutex=0x1d32cb0, md5=..., ignore_timeout=false) at HostDB.cc:668
> #3 0x00000000005e2b34 in HostDBProcessor::getby (this=<value optimized out>,
> cont=0x2b514cc749d0, hostname=0x0, len=<value optimized out>,
> ip=0x2b50e8f092b0, aforce_dns=false, host_res_style=HOST_RES_NONE,
> dns_lookup_timeout=0)
> at HostDB.cc:772
> #4 0x0000000000517f2c in getbyaddr_re (this=0x2b514cc749d0) at
> ../../iocore/hostdb/I_HostDBProcessor.h:417
> #5 HttpSM::do_hostdb_reverse_lookup (this=0x2b514cc749d0) at HttpSM.cc:3968
> #6 0x000000000052f028 in HttpSM::set_next_state (this=0x2b514cc749d0) at
> HttpSM.cc:6932
> #7 0x0000000000518242 in HttpSM::do_hostdb_lookup (this=0x2b514cc749d0) at
> HttpSM.cc:3950
> #8 0x000000000052f44a in HttpSM::set_next_state (this=0x2b514cc749d0) at
> HttpSM.cc:6925
> #9 0x00000000005284fa in HttpSM::handle_api_return (this=0x2b514cc749d0) at
> HttpSM.cc:1559
> #10 0x000000000052ea9a in HttpSM::set_next_state (this=0x2b514cc749d0) at
> HttpSM.cc:6825
> #11 0x000000000052ea8a in HttpSM::set_next_state (this=0x2b514cc749d0) at
> HttpSM.cc:7224
> #12 0x00000000005284fa in HttpSM::handle_api_return (this=0x2b514cc749d0) at
> HttpSM.cc:1559
> #13 0x000000000052ea9a in HttpSM::set_next_state (this=0x2b514cc749d0) at
> HttpSM.cc:6825
> #14 0x00000000005284fa in HttpSM::handle_api_return (this=0x2b514cc749d0) at
> HttpSM.cc:1559
> #15 0x000000000052ea9a in HttpSM::set_next_state (this=0x2b514cc749d0) at
> HttpSM.cc:6825
> #16 0x000000000052fef6 in HttpSM::state_read_client_request_header
> (this=0x2b514cc749d0, event=100, data=<value optimized out>) at HttpSM.cc:821
> #17 0x000000000052a5b8 in HttpSM::main_handler (this=0x2b514cc749d0,
> event=100, data=0x2b514802ca08) at HttpSM.cc:2539
> #18 0x000000000068793b in handleEvent (event=<value optimized out>,
> vc=0x2b514802c900) at ../../iocore/eventsystem/I_Continuation.h:146
> #19 read_signal_and_update (event=<value optimized out>, vc=0x2b514802c900)
> at UnixNetVConnection.cc:138
> #20 0x0000000000689ec4 in read_from_net (nh=0x2b50e2e17c10,
> vc=0x2b514802c900, thread=<value optimized out>) at UnixNetVConnection.cc:320
> #21 0x000000000067fb12 in NetHandler::mainNetEvent (this=0x2b50e2e17c10,
> event=<value optimized out>, e=<value optimized out>) at UnixNet.cc:384
> #22 0x00000000006ac8cf in handleEvent (this=0x2b50e2e14010, e=0x1a9ef30,
> calling_code=5) at I_Continuation.h:146
> #23 EThread::process_event (this=0x2b50e2e14010, e=0x1a9ef30, calling_code=5)
> at UnixEThread.cc:145
> #24 0x00000000006ad273 in EThread::execute (this=0x2b50e2e14010) at
> UnixEThread.cc:269
> #25 0x00000000006abc2a in spawn_thread_internal (a=0x198f820) at Thread.cc:88
> #26 0x00002b50e026b9d1 in start_thread () from /lib64/libpthread.so.0
> #27 0x000000381b2e8b6d in clone () from /lib64/libc.so.6
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
