PM98521: POTENTIAL STORAGE EXHAUSTION IN SYSLOGD

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• If the rules in the /etc/syslog.conf file specify multiple
  remote server names or IP addresses as the destination for
  messages, the potential exists for SYSLOGD to get into a
  recursive call loop if an error occurs during the process of
  sending a logged message.  This will eventually exhaust the
  address space's storage, leading to an ABENDS0878 with reason
  code 00000010.
  
  
  Symptoms in a dump:
  
  An IP VERBX LEDATA 'TCB(xxxxxxxx) CEEDUMP' report specifying the
  abending TCB will show the following sequence repeated,
  terminating after 250 entries:
  
     loglocal    +00000378              EZASLOGD
     logerror    +00000074              EZASLOGD
     slBadSyscall+0000020A              EZASLOGD
     slg_thread_lock
                 +000001FE              EZASLOGD
     slg_serialize_inet
                 +0000008A              EZASLOGD
     CEEVSSFR    +00000012              CEEBINIT
     fprintlog   +00000676              EZASLOGD
     CEEVSSFR    +00000012              CEEBINIT
     outputMsg   +00000200              EZASLOGD
     logmsg      +00000DC2              EZASLOGD
  
     NOTE:  The occurrences of the CEEVSSFR routine depend on the
            current LE runtime options.
  
  The RSAs for the loglocal routine will have the following
  message in them:
  
     syslogd: FSUM1223 Error 74/xxxxxxxx was returned from
     pthread_mutex_lock(inetLock (xxxxxxxx)): EDC5116I Resource
     deadlock avoided.
  
  
  Keywords:
  
  S878  U4088/000003EC
  

Local fix

• Add the following condition to any rule that would match the
  SYSLOGD job and the daemon facility and has a remote address
  reference (use the relevant job name):
  
    *.SYSLOGD.daemon.none
  

Problem summary

• ****************************************************************
  * USERS AFFECTED:                                              *
  * All users of the IBM Communications Server                   *
  * for z/OS Version 2 Release 1 IP: Syslogd                     *
  ****************************************************************
  * PROBLEM DESCRIPTION:                                         *
  * A loop occurred after an error                               *
  * sending a message to a remote                                *
  * syslogd.  Storage was filled with                            *
  * messages and save areas indicating                           *
  * that the attempt to send the error                           *
  ****************************************************************
  * RECOMMENDATION:                                              *
  * Apply PTF                                                    *
  ****************************************************************
  The error path for sending an error message that had the INET
  lock held while attempting to send the error message to
  another remote syslogd failed when syslogd went into a loop.
  +-------------------------------------------------------------+
  + Please check our Communications Server for z/OS homepages   +
  + for common networking tips and fixes.  The URL for these    +
  + homepages can be found in Informational APAR II11334.       +
  +-------------------------------------------------------------+
  

Problem conclusion

• SYSLOGD is amended to release the INET lock before sending the
  error message.
  

Temporary fix

Comments

APAR Information

• APAR is sysrouted FROM one or more of the following:

  PM94151

• APAR is sysrouted TO one or more of the following:

  UK98367

Modules/Macros

• EZASLOGD EZASYARC EZASYCMD EZASYUTL EZASYTHD
  

Fix information

• Fixed component name

  TCP/IP V3 MVS

• Fixed component ID

  5655HAL00

Applicable component levels

• R210 PSY UK98367

     UP13/11/19 P F311

Fix is available

• Select the PTF appropriate for your component level. You will be required to sign in. Distribution on physical media is not available in all countries.