Pivotal Knowledge Base


How to do Root Cause Analysis for the GPText Instance Down Issue


  • Pivotal Greenplum Database (GPDB) 4.3.x
  • Operating System- Red Hat Enterprise Linux 6.x
  • GPText 1.3 and 2.0


What is broken or in other words "Customer Impact"?


<Information to be collected for this JIRA - if information is not provided, please explain why>

  • GPDB version
    select version();
  • GPText version
    select * from gptext.version();
  • GPDB segment status
    select * from gp_segment_configuration order by content;
    select * from gp_configuration_history order by 1 desc ;
  • GPText instance status
  • GPText health check
    gptext-state --healthcheck
  • gptext-detective output or collect below manually:
    GPDB master log
    GPDB segment log (where the GPText instance goes down)
    GPText solr log (where the GPText instance goes down)
    gpAdminLogs/gptext-*.log on master.
  • gp_interconnect_type GUC value
    gpconfig -s gp_interconnect_type
  • GPDB memory setting
    gpconfig -s gp_vmem_protect_limit
  • Segment server physical memory size
  • gptext related GPDB GUCs on master:
    cat postgresql.conf|grep gptext
  • On problematic segment server:
    netstat -anp
    netstat -i
    ps -ef|grep solr|grep -v grep



Powered by Zendesk