Pivotal Knowledge Base

Follow

How to install and use gpsupport

Environment

Product Version
Pivotal Greenplum (GPDB) All versions

Purpose

Pivotal Support generally requires Pivotal Greenplum database log files in order to diagnose issues. It is recommended to provide at least the master logs when opening a ticket. This will allow Pivotal Support to get to work more quickly and understand the issue.

Cause

gpsupport is the recommended tool for collecting logs, this will collect the most common logs from master and segment servers and generate a single archive file. This archive can be uploaded to Pivotal Support.

This articles provides instructions to download and install gpsupport as well as some common scenarios.

Additional information can be found in the gpsupport documentation.

Resolution

Follow the instructions to download

  1. Go to Pivotal Greenplum Database on Pivotal Network.
  2. Go to Greenplum Support.
  3. Download the latest release of Greenplum Database Support Utility.
  4. Copy the file to the GPDB master node (mdw).

Follow the instructions to Install

  1. From the GPDB master node (mdw), extract the file:
    gunzip gpsupport-X.X.X.X.gz
  2. Make the file executable:
    chmod +x gpsupport-X.X.X.X

Common scenarios

  • Collect logs from segment from current date:
$ ./gpsupport collect logs
Checking connectivity and authentication...
Validating segment file on remote hosts...
Segment file not found or invalid on some hosts. Installing...
Installation complete.
Starting node collection
Checking for errors in node collection
Generating final tarfile /tmp/log_collector_2015-08-13_06-11-21-000.tar
  • Collect logs from the last 3 days:
$ ./gpsupport startDate $(date '+%Y-%m-%d' --date='3 days ago') endDate $(date '+%Y-%m-%d') collect logs
  • Diagnose UDP connectivity between segment hosts.
  • Note that the mandatory hostfile used in conjunction with the tool should have a single hostname on each line and no extraneous whitespace.
$ ./gpsupport-1.2.0.0 hostfile=/home/gpadmin/hostfile diagnose connectivity
Checking connectivity and authentication...
Validating segment file on remote hosts...
~# Starting remote senders
smdw --> ALL NODES
sdw3 --> ALL NODES
mdw --> ALL NODES
sdw5 --> ALL NODES
~# Waiting for remote senders to start
~# Starting workload
~# Waiting for workload to complete
~# Completed workload
~# Stopping remote senders
~# IO Stream report
smdw --> ALL NODES | 11.41mb/s | 0.00000% Loss | 342mb sent | 342mb received
        --> sdw3:7114 | 3.80mb/s | 0.00000% Loss | 114mb sent | 114mb received
        --> mdw:7114 | 3.80mb/s | 0.00000% Loss | 114mb sent | 114mb received
        --> sdw5:7114 | 3.80mb/s | 0.00000% Loss | 114mb sent | 114mb received
sdw3 --> ALL NODES | 20.73mb/s | 4.44175% Loss | 650mb sent | 622mb received
        --> mdw:7115 | 7.23mb/s | 0.00000% Loss | 216mb sent | 216mb received
        --> sdw5:7115 | 7.23mb/s | 0.00000% Loss | 216mb sent | 216mb received
        --> smdw:7115 | 6.27mb/s | 13.32493% Loss | 216mb sent | 188mb received
mdw --> ALL NODES | 20.85mb/s | 5.28495% Loss | 660mb sent | 625mb received
        --> smdw:7116 | 6.18mb/s | 15.85465% Loss | 220mb sent | 185mb received
        --> sdw3:7116 | 7.34mb/s | 0.00000% Loss | 220mb sent | 220mb received
        --> sdw5:7116 | 7.34mb/s | 0.00000% Loss | 220mb sent | 220mb received
sdw5 --> ALL NODES | 20.71mb/s | 4.53547% Loss | 650mb sent | 621mb received
        --> smdw:7117 | 6.25mb/s | 13.60608% Loss | 216mb sent | 187mb received
        --> sdw3:7117 | 7.23mb/s | 0.00000% Loss | 216mb sent | 216mb received
        --> mdw:7117 | 7.23mb/s | 0.00000% Loss | 216mb sent | 216mb received
~# Finished

Comments

Powered by Zendesk