[
https://issues.apache.org/jira/browse/TS-716?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13024824#comment-13024824
]
Kissdev commented on TS-716:
----------------------------
Under the same load and configuration, the new testing shows the trunk does not
crash,but throws the following errors after running for serveral minutes
(leading to ERR_CLIENT_ABORT and ERR_CONNECT_FAIL response for all requests):
[Apr 25 23:35:07.867] Manager {46969923497680} NOTE: [LocalManager::startProxy]
Launching ts process
Layout configuration
--prefix = '/ats'
--exec_prefix = '/ats'
--bindir = '/ats/bin'
--sbindir = '/ats/bin'
--sysconfdir = '/ats/etc/trafficserver'
--datadir = '/ats/share/trafficserver'
--includedir = '/ats/include'
--libdir = '/ats/lib'
--libexecdir = '/ats/libexec/trafficserver'
--localstatedir = '/ats/var'
--runtimedir = '/ats/var/trafficserver'
--logdir = '/ats/var/log/trafficserver'
--mandir = '/ats/man'
--infodir = '/ats/info'
--cachedir = '/ats/var/trafficserver'
[TrafficServer] using root directory '/ats'
[Apr 25 23:35:07.873] Manager {46969923497680} NOTE:
[LocalManager::pollMgmtProcessServer] New process connecting fd '8'
[Apr 25 23:35:07.873] Manager {46969923497680} NOTE: [Alarms::signalAlarm]
Server Process born
[Apr 25 23:35:08.884] {47969045663584} STATUS: opened
/ats/var/log/trafficserver/diags.log
[Apr 25 23:35:08.884] {47969045663584} NOTE: updated diags config
[Apr 25 23:35:08.887] Server {47969045663584} NOTE: cache clustering disabled
[Apr 25 23:35:08.912] Server {47969045663584} NOTE: cache clustering disabled
[Apr 25 23:35:09.056] Server {47969045663584} NOTE: logging initialized[7],
logging_mode = 3
[Apr 25 23:35:09.068] Server {47969045663584} NOTE: traffic server running
[Apr 25 23:35:15.375] Server {1132697920} NOTE: cache enabled
FATAL: MultiCache.cc:1343: failed assert `!"out of space"`
/ats/bin/traffic_server - STACK TRACE:
/ats/lib/libtsutil.so.2(ink_fatal_va+0xb6)[0x2ba0a9359306]
/ats/lib/libtsutil.so.2(ink_fatal+0xe4)[0x2ba0a93594de]
/ats/lib/libtsutil.so.2(_ink_assert+0xaa)[0x2ba0a935783e]
/ats/bin/traffic_server(_ZN14MultiCacheBase5allocEPii+0xa4)[0x638e98]
/ats/bin/traffic_server(_ZN18HostDBContinuation8dnsEventEiP7HostEnt+0x944)[0x631424]
/ats/bin/traffic_server(_ZN12Continuation11handleEventEiPv+0x6f)[0x4d25ff]
/ats/bin/traffic_server(_ZN8DNSEntry9postEventEiP5Event+0xda)[0x63eea8]
/ats/bin/traffic_server(_ZN12Continuation11handleEventEiPv+0x6f)[0x4d25ff]
/ats/bin/traffic_server(_ZN7EThread13process_eventEP5Eventi+0x11e)[0x6f4184]
/ats/bin/traffic_server(_ZN7EThread7executeEv+0x94)[0x6f4376]
/ats/bin/traffic_server[0x6f3ad6]
/lib64/libpthread.so.0[0x35808064a7]
/lib64/libc.so.6(clone+0x6d)[0x357fcd3c2d]
> Crash in Continuation::handleEvent
> -----------------------------------
>
> Key: TS-716
> URL: https://issues.apache.org/jira/browse/TS-716
> Project: Traffic Server
> Issue Type: Bug
> Components: Core
> Affects Versions: 2.1.7
> Environment: CentOS 5.4 x86_64, 6 * 2T SATA Disks, 48G Memory
> Reporter: Kissdev
> Assignee: John Plevyak
> Priority: Critical
> Fix For: 2.1.8
>
> Attachments: crasher.patch
>
>
> ATS crashes with the following configuration:
> - reverse proxy , storage: 6 raw devices (6*2T), 1 partition (2T)
> - remap config:regex_map http://(.*) http://$1
> The load : about 100Mbps, requests for top 4000 internet sites, mainly
> html,js,pictures,flashes....
> Detail of crashes by core dump:
> crash #1:
> {code}
> #0 0x00000000004dd17a in Continuation::handleEvent (this=0x2aaaba364cb0,
> event=1, data=0x90f7170) at I_Continuation.h:146
> 146 I_Continuation.h: No such file or directory.
> in I_Continuation.h
> (gdb) bt
> #0 0x00000000004dd17a in Continuation::handleEvent (this=0x2aaaba364cb0,
> event=1, data=0x90f7170) at I_Continuation.h:146
> #1 0x0000000000702b80 in EThread::process_event (this=0x2aaaab101010,
> e=0x90f7170, calling_code=1) at UnixEThread.cc:140
> #2 0x0000000000702fa1 in EThread::execute (this=0x2aaaab101010) at
> UnixEThread.cc:232
> #3 0x00000000007024d2 in spawn_thread_internal (a=0x8d94a70) at Thread.cc:85
> #4 0x00000036ebc064a7 in start_thread () from /lib64/libpthread.so.0
> #5 0x00000036eb0d3c2d in clone () from /lib64/libc.so.6
> (gdb) frame 0
> #0 0x00000000004dd17a in Continuation::handleEvent (this=0x2aaaba364cb0,
> event=1, data=0x90f7170) at I_Continuation.h:146
> 146 in I_Continuation.h
> (gdb) print *this
> $1 = {<force_VFPT_to_top> = {_vptr.force_VFPT_to_top = 0x2aaaba360a11},
> handler = &virtual table offset -1157442765409226770, this adjustment
> -1157442765409226769,
> handler_name = 0xefefefefefefefef <Address 0xefefefefefefefef out of
> bounds>, mutex = {m_ptr = 0xefefefefefefefef}, link = {<SLink<Continuation>>
> = {
> next = 0xefefefefefefefef}, prev = 0xefefefefefefefef}}
> {code}
> crash #2:
> {code}
> (gdb) bt
> #0 0x00000000004dd17a in Continuation::handleEvent (this=0x2aaabc0bce80,
> event=1, data=0x154b5a80) at I_Continuation.h:146
> #1 0x00000000006db290 in InactivityCop::check_inactivity (this=0x154c8730,
> event=2, e=0x154b5a80) at UnixNet.cc:57
> #2 0x00000000004dd1bb in Continuation::handleEvent (this=0x154c8730,
> event=2, data=0x154b5a80) at I_Continuation.h:146
> #3 0x0000000000702b80 in EThread::process_event (this=0x2aaaab606010,
> e=0x154b5a80, calling_code=2) at UnixEThread.cc:140
> #4 0x0000000000702ec2 in EThread::execute (this=0x2aaaab606010) at
> UnixEThread.cc:217
> #5 0x00000000007024d2 in spawn_thread_internal (a=0x154852c0) at Thread.cc:85
> #6 0x00000036ebc064a7 in start_thread () from /lib64/libpthread.so.0
> #7 0x00000036eb0d3c2d in clone () from /lib64/libc.so.6
> (gdb) frame 0
> #0 0x00000000004dd17a in Continuation::handleEvent (this=0x2aaabc0bce80,
> event=1, data=0x154b5a80) at I_Continuation.h:146
> 146 in I_Continuation.h
> (gdb) print *this
> $1 = {<force_VFPT_to_top> = {_vptr.force_VFPT_to_top = 0x16280061},
> handler = &virtual table offset -1157442765409226770, this adjustment
> -1157442765409226769,
> handler_name = 0xefefefefefefefef <Address 0xefefefefefefefef out of
> bounds>, mutex = {m_ptr = 0xefefefefefefefef}, link = {<SLink<Continuation>>
> = {
> next = 0xefefefefefefefef}, prev = 0xefefefefefefefef}}
> (gdb)
> {code}
> crash #3:
> {code}
> (gdb) bt
> #0 0x00000000004dd17a in Continuation::handleEvent (this=0x2aaab45d3a10,
> event=2, data=0x5631120) at I_Continuation.h:146
> #1 0x0000000000702b80 in EThread::process_event (this=0x2aaaaabfc010,
> e=0x5631120, calling_code=2) at UnixEThread.cc:140
> #2 0x0000000000702ec2 in EThread::execute (this=0x2aaaaabfc010) at
> UnixEThread.cc:217
> #3 0x000000000050917c in main (argc=3, argv=0x7fff0af6e3b8) at Main.cc:1962
> (gdb) frame 0
> #0 0x00000000004dd17a in Continuation::handleEvent (this=0x2aaab45d3a10,
> event=2, data=0x5631120) at I_Continuation.h:146
> 146 in I_Continuation.h
> (gdb) print *this
> $1 = {<force_VFPT_to_top> = {_vptr.force_VFPT_to_top = 0x2aaab45df291},
> handler = &virtual table offset -1157442765409226770, this adjustment
> -1157442765409226769,
> handler_name = 0xefefefefefefefef <Address 0xefefefefefefefef out of
> bounds>, mutex = {m_ptr = 0xefefefefefefefef}, link = {<SLink<Continuation>>
> = {
> next = 0xefefefefefefefef}, prev = 0xefefefefefefefef}}
> {code}
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
