*** trafodion_scanner.cfg ***
+ Extended the specification format for configuration checks: added a new
<severity level> field to differentiate between required and recommended
configuration items. The possible values are 'error' and 'warning'. Added
appropriate values for the existing configuration checks.
+ Fixed configuration gathering and check commands for "GB of total physical
memory" to round the GB value "up" rather than "down".
This is needed because systems often have close to 16GB of memory, but not
exactly that much, e.g., VMs that Chris Sheedy uses.
+ Added the following configuration gathering commands:
1. print the Hadoop distribution type (based on <<<hadoop_distribution>>>)
2. print the Hadoop version
3. print the HBase version
4. print the HBase Java version
+ Added the following configuration checks:
1. check for supported HBase version
2. check for supported HBase Java version
3. check for recommended unixODBC version
4. check for that Hyperthreading is enabled, if the capability exists
+ Fixed the ValidNodeName configuration check to detect all possible special
characters that are not supported, although more work is needed to
syncronize the checking logic with the installer
+ In the explanation comments for the configuration check specification
format, clarified how the <comparison operator> is applied.
+ Improved the wording and capitalization in the descriptions of the
configuration gathering and check commands.
+ Updated Copyright for 2015
*** trafodion_scanner ***
+ Added code for processing the new <severity level> field and reporting
it in the results output
+ Added checks for the Hadoop distribution. Errors will be generated for
the following cases (program will exit):
1. a node has no Hadoop distribution installed (Ambari or Cloudera)
2. a node has multiple Hadoop distribution installed (Ambari and Cloudera)
3. the set of nodes has mixed Hadoop distributions (some nodes have Ambari
and some have Cloudera)
4. Hadoop distribution cannot be determined due to some runtime error
Also, added the ability to obtain the Hadoop distribution type via a special
placeholder variable that can be specified in the <command> string for
configuration gathering and check commands: <<<hadoop_distribution>>>,
with possible values 'ambari' and 'cloudera'.
+ Added a special mode for Chris Sheedy, to allow use of the scanner on
workstations and VMs without requiring sudo access, passwordless ssh,
and a Hadoop distribution.
+ Cleaned up the source code to make it more readable,
e.g., tried to avoid very long lines, etc
+ Improved log file output, e.g., added expected result information
+ Made all of the exit codes consistent, per the following conditions:
0 ~ Success
1 ~ One or more errors encountered (invalid arguments,
prerequisites not satisfied, can't execute check command, etc)
2 ~ At least one check failed with error severity (required configuration)
3 ~ At least one check failed with warning severity
(recommended configuration) and no checks failed with error severity
+ Added --log_file option to specify the log filename
+ Improved error handling
+ Added logic to verify that the --nodes (-n) command-line argument
does not contain IP addresses, although more work is needed to
syncronize the argument checking with the installer
+ Updated Copyright for 2015
Change-Id: I6774adc7c39a00d17fb7a93f82e24fbb03c3aa7e
2 files changed