System information

Resolved Issues in Greenplum Database 4.2.7.3 4
Greenplum Database 4.2.7.3 Release Notes
22060 Functions and
Languages,
Resource
Management
4.2.7.1 The string_to_array function fails to produce an array when the input
string is larger than 250MB.
22045 Dispatch 4.2.7.1 Reporting errors when the maximum number of connections on
Greenplum Database segment instances were exceeded has been
enhanced.
• Error message for failed connections to segment instances are
printed to pg_log without enabling debug mode.
• The error message displayed on the client is
segworker group creation failed.
21880 Dispatch 4.2.7.1 An internal error message was displayed for an error that occurred for
an unsupported query plan with multiple writer segworker groups. The
message has been enhanced and provides a description and more
information about the context. This is the updated error message.
ERROR: Unsupported query plan with multiple writer gangs.
HINT: Likely caused by a user defined function that reads or modifies
data in a distributed table.
CONTEXT: SQL statement "SQL_statement"
21795 Fault Detection
and
Transitions
4.2.7.1 The Greenplum Database server configuration parameter gp_log_fts
specifies the level of fault probing messages written to the log file. Valid
values are off, terse, verbose, and debug. The verbose mode provides
minimal useful log messages.
After the value of gp_log_fts has been changed, stop and restart
Greenplum Database with the command gpstop -u for the value to take
affect.
21694 Query
Execution
4.2.7.1 When a Hash Join or Hash Aggregate operator spilled to disk for an
SQL query, the amount of workfile bytes read and written to disk was
incorrectly reported in the EXPLAIN ANALYZE output.
21630
19106
Resource
Management
4.2.7.1 Dropping the resource queue used by an active session from another
session would result in a postmaster reset on the master. This has
been changed. Now, instead of a reset, an error message is generated
in the affected session.
21545 Access
Methods
4.2.7.1 On some occasions when a pg_terminate_backend command was
issued, Greenplum Database encounters a PANIC and performs a
recovery.
21050 Monitoring:
gpperfmon
server,
Monitoring:
gpperfmon UI
4.2.7.1 Under some conditions, the Greenplum Command Center process
gpsmon would consume a large amount of memory.
20997 Monitoring:
gpperfmon
server,
Monitoring:
gpperfmon UI
4.2.7.1 The performance of Greenplum Command Center was slow when
processing log alerts from a very large log file.
The log processing has been enhanced and the Greenplum Database
server configuration parameter gpperfmon_log_alert_level can control
the growth of the gpperfmon log.
20970 Query
Execution
4.2.7.1 When Greenplum Command Center had been running for a long time,
the query it used for disk usage information could not be stopped.
Table 1 Resolved Issues in 4.2.7.3
Issue
Number
Category
Resolved
in
Description