[jira] [Created] (TS-2305) trafficserver restarts with !t_state.host_db_info.reverse_dns assert

["Sjaak Westdijk (JIRA)" <ji...@apache.org>]

Sjaak Westdijk created TS-2305:
----------------------------------

             Summary: trafficserver restarts with !t_state.host_db_info.reverse_dns assert
                 Key: TS-2305
                 URL: https://issues.apache.org/jira/browse/TS-2305
             Project: Traffic Server
          Issue Type: Bug
          Components: DNS
            Reporter: Sjaak Westdijk


Running ATS 4.0.1 on OmniOS we observing a couple trafffic_server restarts on busy moments. At that time cpu load peakes and the system client throughput peakes to 1000Mbits. The messages in the system log are:

 FATAL: HttpSM.cc:2048: failed assert `!t_state.host_db_info.reverse_dns`
 and
 FATAL: HttpSM.cc:2049: failed assert`ats_is_ip(t_state.host_db_info.ip())`

On the restart this in the diag.log
{noformat}
[Oct 29 12:03:21.892] Server {0x1} NOTE: cache clustering disabled
[Oct 29 12:03:21.954] Server {0x1} NOTE: ip_allow.config updated, reloading
[Oct 29 12:03:21.980] Server {0x1} WARNING: unable to set file '/opt/ts/var/trafficserver/host.db' size to 25935872: 22, Invalid argument
[Oct 29 12:03:21.980] Server {0x1} WARNING: header missing/corrupt: [hostdb.config] : reinitializing database
[Oct 29 12:03:21.980] Server {0x1} NOTE: reconfiguring host database
[Oct 29 12:03:21.981] Server {0x1} WARNING: unable to set file '/opt/ts/var/trafficserver/host.db' size to 25935872: 22, Invalid argument
[Oct 29 12:03:22.019] Server {0x1} NOTE: cache clustering disabled
[Oct 29 12:03:22.021] Server {0x1} NOTE: logging initialized[15], logging_mode = 3
[Oct 29 12:05:10.011] Server {0x1} NOTE: traffic server running
[Oct 29 12:05:26.915] Server {0x6} NOTE: cache enabled
{noformat}

This in the manager.log
{noformat}
[Oct 29 12:03:19.711] Manager {0x1} ERROR: [LocalManager::pollMgmtProcessServer] Server Process terminated due to Sig 6: Abort
[Oct 29 12:03:19.718] Manager {0x1} ERROR:  (last system error 2: No such file or directory)
[Oct 29 12:03:19.718] Manager {0x1} ERROR: [Alarms::signalAlarm] Server Process was reset
[Oct 29 12:03:19.718] Manager {0x1} ERROR:  (last system error 2: No such file or directory)
[Oct 29 12:03:20.724] Manager {0x1} NOTE: [LocalManager::startProxy] Launching ts process
[Oct 29 12:03:20.745] Manager {0x1} NOTE: [LocalManager::pollMgmtProcessServer] New process connecting fd '12'
[Oct 29 12:03:20.745] Manager {0x1} NOTE: [Alarms::signalAlarm] Server Process born
{noformat}

This is the dns config in records.config
{noformat}
CONFIG proxy.config.dns.search_default_domains INT 0
CONFIG proxy.config.dns.splitDNS.enabled INT 0
CONFIG proxy.config.dns.max_dns_in_flight INT 2048
   # Additional URL expansions for http DNS lookup
CONFIG proxy.config.dns.url_expansions STRING NULL
CONFIG proxy.config.dns.round_robin_nameservers INT 0
CONFIG proxy.config.dns.nameservers STRING 127.0.0.1
CONFIG proxy.config.dns.resolv_conf STRING /etc/resolv.conf
   # This provides additional resilience against DNS forgery, particularly in
   # forward or transparent proxies, but requires that the resolver populates
   # the queries section of the response properly.
CONFIG proxy.config.dns.validate_query_name INT 0
{noformat}

We are running a local dns server which is actually a forwarding server with a blockingzone for botnet sites.

Note: the host.db file keeps its 0 bytes size all the time
 



--
This message was sent by Atlassian JIRA
(v6.1#6144)