IBM DataPower Operations Dashboard v1.0.10.0

A newer version of this product documentation is available.

You are viewing an older version. View latest at IBM DPOD Documentation.

Store

The Store screen is accessible by logging into the Web Console and navigating to [Manage→Internal Health→Store].

DPOD's Big Data Store is where all of DPOD's data is stored. 

This screen display consist of 5 widgets, each allowing access to view data or perform Store-related tasks. They are described below and in the following pages:

Health Data

The top 3 widgets display health data related to the store.

Cluster Status

DetailDescriptionDesired State
ex-raw-trans

The general state of the Big Data store.

  • GREEN – Everything is working properly.
  • YELLOW – System is up and running, but there’s a problem requiring your attention (e.g. one of the nodes is down, but another node took over)
  • RED – There is a major problem with the Big Data store. DPOD functions are offline. Contact support.

Note: During the first 10-20 minutes following a system startup, a value of YELLOW or RED is normal.

GREEN

(but RED and YELLOW acceptable during startup)

 

Shards

Details of the allocation of logical data to physical storage in the Big Data Store.
All shards should be assigned, and unassigned shards number should be 0.

Note: During the first 10-20 minutes following a system startup, a value greater than zero is normal. 

If unassigned shards persist, see Unassigned Shards under troubleshooting.

0 Unassigned shards.
(but a larger than 0 value acceptable during startup) 
ButtonsTwo buttons are available to the far right of the cluster status widget. Both the ES-head and ES-Visualize buttons are for DPOD support troubleshooting tasks. When required, DPOD support will guide you through the usage of these buttons. 

Index Sets

Data sets that store monitoring data received from monitored devices. 

DetailDescriptionDesired State
NameIndex namen/a
Indices Count

Indices counts should be about 1 per day or at most 2 per day.
The number of days stored in an index set can be derived by finding the delta in days between the oldest and newest document time stored inside it.
The indices count should therefore be no more than twice that value.

When the Indices count is higher, query performance may be affected. See Increase Storage for Index Sets for ways to resolve this.sue.

At most - twice the number
of  days stored in data set.
SizeSize of the data set.Each index should contains data between 1 day to 2 maximum
Documents Count

All data sets starting with “wdp-“ should have a document count larger than 0.

A value of 0 for a wdp-* data set points to problems receiving data from monitored devices.

> 0
Oldest Document Time

Oldest record in the index set. Verify that this fits your requirements, and if not – you may need to Increase Storage for Index Sets.

 

As per your
requirements 
Newest Document Time

Newest record in the index set. Verify that this fits your requirements, and if not – you may need to Increase Storage for Index Sets.

 

As per your
requirements

DPOD uses index sets to manage different data types in the Big Data store.

Index set NameDescription

wdp-syslog

Contains log records collected by the Syslog agents from monitored devices, except system messages – e.g. transaction started, network call states, integration code errors, etc.

wdp-syslog-sys-*

Contains log records of system events collected by the Syslog agent from monitored devices.

wdp-wsm

Contains payload message collected by the WS-M agent when WS-M recording is enabled.

wdp-device-resources

Contains data collected by the device resources agent, e.g. monitored devices CPU, memory, hardware state, etc.

wdp-service-resources

Contains data collected by the service resources agent, e.g. memory consumption by monitored services, etc.

Other index sets are restricted and are for internal use only.

These indexes are self-managed from a disk storage perspective. Each index set is divided into multiple indexes and when the allocated space for that index set is exhausted, the oldest index set is deleted and new one is created.
This operation is done by the retention housekeeping process.

Nodes

These are the data nodes which the store uses for its internal processing.

DetailDescriptionDesired State
NameNode Namen/a
VersionNode Versionn/a
Host NameNode's host namen/a
Type

A correctly set up system should have at least one ‘D’ node (data processing node),
and an odd number of ‘M’ nodes (master nodes).

  • D Nodes > 0
  • Odd M Nodes number 

Nodes Health

The Nodes Health table contains additional information about the ElasticSearch nodes (for example, the amount of swap space used, average time to index documents, etc)
Hover over the data to get additional information.
 

IBM DataPower Operations Dashboard (DPOD) v1.0.10.0