Summary for Patch 2170 (Verified)

Summary Improvement for server processing
Platform Unix
Category bugfix
Valid for 6.4.x, 6.5.0.x, 6.5.1.14
Requires 2137, 6.4.0.9 - 6.5.0.22: 2108, 6.4.0.9 - 6.5.0.22: 2096, 6.4.0.9 - 6.5.0.22: 2089
Client Required Patch/es
Supersedes 2118, 2104, 2056, 2025, 2023, 2017, 2016, 917, 857
Superseded by 2179 : Released / Hidden
Affects Server: Core
Last Updated January 9, 2019

Download now

Download_red patch2170.tar.gz

Description

This patch:
- fixes the problem with "add --redo"
- fixes the problem with overloaded server because of check_and_split_data
- adds a directive in openit.cfg, "COMPRESSION", which will default to "global=none,temp=none". By
setting the temp to "gzip", and restarting the server, temporary files in the server will be
compressed with gz format.
- modifies Archiver to handle .gz incoming files besides of plain .in files
- makes a directory for incoming/archiver and moves archiver files into it to speed up processing of
check_and_split_data to avoid overloading of server.
- fixes issue with check_and_split_data failing to process files with timed out lock
- fixes issue with check_and_split_data hanging if lock files are accumulated in the incoming
directory

The following perl modules, scripts, and binaries are updated:
* Config.pm
* DataLoader.pm
* DataTypes.pm
* Defaults.pm
* File.pm
* FileHelper.pm
* FileLock.pm
* FileObject.pm
* Server.pm
* Protocol.pm
* add
* check_and_split_data
* merge_raw_data
* filtrator
* archiver
* calcchecksum
* summasummarum

merge_psm_data is added to support compression of temporary data in the server.

Supersedes Patch Description

Expand all

Patch 2118 — Fix for warning messages that are not recognized by Filtrator
This patch updates Filtrator to fix the warning messages that are not recognized.
Patch 2104 — New mapping type 'Advance Mapping'
This patch contains a new feature called ‘Advance Mapping’ which can be configured to
include-exclude data like products and features. This feature can be utilized when merging data and
excluding the unwanted data.

Usage:
The configuration is the same as a normal mapping but prefixes should be added to indicate if it’s
to be included or excluded. Adding ‘+’ or no prefixes on a name means it is included, adding ‘-’ on
a name means it is excluded. Entries are evaluated from bottom-up.

The mapping type should also be set as 'ADVMAP'.

Example:

 	ADVMAP: Combined_Product
   	+*;licsrv
   	-server01_legacy;licsrv

This means all products that match(es) ‘*;licsrv’ will be included and 'server01_legacy;licsrv' will
be excluded in the mapping.

NOTE:

- Prefixes mentioned cannot be used for regular MAP setup.
- Advance Mapping is not recommended for rename functions.
- While ADVMAP enables a more advanced method for mapping, it is also slower. It can be significant
for huge data sets.
Patch 2056 — Fix for periodic reports on Unix Core Server
General Description: A bug in Unix Core Server can cause missing periodic reports under certain
circumstances.

Detailed Description: When the Unix Core Server is configured with WEEKLY_REPORT_WAIT_DAYS, it may
not work if the period to be reported on is in the previous year. It can only happen in the
beginning of January. This patch has a fix for Server.pm to solve this problem.
Patch 2025 — Handling of duplicate data and Daylight Saving Time switch
General Description:

This patch removes duplication of data when licpoll has collected overlapping
files and fix for missing hours after Daylight Saving Time switch.

Detailed Description:

When two or more files was collected from licpoll having the same product,
interval and offset for the same collection period, the first sample would be
duplicated, giving too high values for certain measurements, e.g. max in use.
This has now been compensated for by filtrator, by selecting only one licpoll
file for each interval, offset and product whenever there is samples in the
beginning of the hour.

Report shows missing hours in a 24 hour report. This bug is cause by the recent daylight saving time
switch. This patch will update data processing to handle the DST switch issue.
Note: You may have to regenerate the data that has been affected by the DST issue.

Patch 2023 — Update of data and archive processing to match jobscheduler timing
When the jobscheduler was introduced, the timing of the other backend processes
needs to be updated. The archiving is now completely decoupled from the database
processing.

This is done in the Unix server daemon with this patch.
Patch 2017 — Fix for hostid in RHEL 7
General Description:
This patch contains fix for incomplete output of hostid program in RHEL 7.

Detailed Description:
In RHEL 7, hostid outputs an incomplete id. This patch contains an updated Config.pm which is used
by the hostid program to generate an id. Only the Config.pm file should be updated. Once this patch
has been applied, hostid will now generate a correct host id.
Patch 2016 — Fix for wrong data under hostname field
General Description: Data under host name field shows port number.

Detailed Description: 

The field for host name shows port number when generating licpoll archives with version 2.4. This is
caused by faulty parsing of filtrator. This patch will update and fix this parsing issue.

Note:  You may need to regenerate affected database data. 
Patch 917 — Added support for delay in generating weekly reports
This patch contains the support for delay in generating weekly reports. 

Add the line WEEKLY_REPORT_WAIT_DAYS <delay in days> to /etc/opt/openit.cfg to activate.
Patch 857 — Addition of offset to map maintenance
This is to add option to change offset of map maintainer.

Installation

Unpack tar file in e.g /tmp/patch on the Open iT host. Do the following as root:

	cd /tmp/patch
	./patch.pl

Corresponding Windows Patches

No corresponding Patches available.



Back to Version List