Cisco Virtualized Video Processing Controller User Guide - Release 3.3.5

Alarms, Events, and Statistics

---

This chapter describes the procedures for viewing, backing up, and restoring V2PC statistics. It also lists the alarms and events that V2PC master and worker nodes can generate, and the statistics V2PC can collect.

Introduction

InfluxDB, a database for time series data, aggregates control plane statistics for the V2PC platform and its applications. The InfluxDB database instance runs on the ELK node VM.

All statistics from various V2PC nodes are sent to InfluxDB through the Sensu monitoring platform. The Sensu server runs in a high-availability (HA) configuration on the V2PC master nodes. Through Sensu checks, the statistics handler collects statistics from each node or service and sends it to InfluxDB running on the ELK node.

Note For additional information on InfluxDB, see the documentation for InfluxDB version .12 at: https://docs.influxdata.com/influxdb/v0.12

Figure 12-1 V2PC Statistics Aggregation

The InfluxDB metastore contains internal information about the status of the system, including user information, database and shard metadata, and enabled retention policies. Because there is only one ELK node instance in a V2PC deployment, there is no high availability on the statistics data aggregated in InfluxDB. As a result, the loss of the ELK node or InfluxDB instance can result in loss of this data.

We therefore recommend performing a regular backup schedule for statistics data. The following sections provide instructions for logging in to the InfluxDB shell, performing a backup of the database, and if ever necessary, restoring the database from a previous backup.

Viewing Current InfluxDB Data

To log in to InfluxDB on ELK node and use the database shell to view stored statistics data:

Step 1 On the master node, execute the command consul members to identify the active ELK node.

Step 2 Log in to the ELK node via SSH. The default login credentials are:

• User ID: root
• Password: cisco

Step 3 On the command line, type influx to access the InfluxDB command shell prompt.

Step 4 Type auth admin default.

Step 5 Type show databases.

Step 6 Type use stats_system to access the stats_system database, where V2PC stores statistics data.

Step 7 Type show measurements to show a list of measurements (similar to tables in an SQL database).

For example, COS node statistics are stored in the aic_cosnodestats measurement.

Backing Up InfluxDB Data

If a node is running, you can create a backup of the metastore for an instance by executing the command influxd backup /influxdb_backups/metadata. This command backs up the InfluxDB metadata to the directory /influxdb_backups/metadata.

To back up the InfluxDB database:

Step 1 Log in to the ELK node as described in Viewing Current InfluxDB Data.

Step 2 Run the command influxd backup –database stats_system /influxdb_backups/database to back up the influxDB database stats_system to the directory /influxdb_backups/database.

To perform a backup of a remote node, specify the host and port of the remote instance using the -host configuration switch, as show in the following example:

Restoring InfluxDB Data

Note Restoring from a backup is only supported while the InfluxDB daemon is stopped. To restore from a backup, you must provide the path to the backup.

To restore the InfluxDB database from an earlier backup:

Step 1 Log in to the ELK node as described in Viewing Current InfluxDB Data.

Step 2 Run the command influxd restore /tmp/backup to restore the contents of /tmp/backup.

The following optional flags are available for this command:

• -metadir <path to meta directory> – This is the path to the meta directory to which the metastore backup should be recovered. For packaged installations, specify /var/lib/influxdb/meta.
• -datadir <path to data directory> – This is the path to the data directory to which the database backup should be recovered. For packaged installations, specify /var/lib/influxdb/data.
• -database <database> – This is the database to which the data should be restored. This option is required if no -metadir option is provided.
• -retention <retention policy> – This is the target retention policy to which the stored data should be restored.
• -shard <shard id> – This is the shard data that should be restored. If specified, -database and -retention must also be set.

Step 3 Restore the backup in two steps, as follows:

a. Restore the metastore so that InfluxDB knows which databases exist:

$ influxd restore -metadir /var/lib/influxdb/meta /tmp/backup/<metastore>

Where <metastore> is the metastore to be restored. For example, to restore meta.00:

b. Recover the backed up stats_system database as follows:

$ influxd restore -database stats_system -datadir /var/lib/influxdb/data /tmp/backup

Sample Queries

Use the following queries from the database shell to show database, measurement, tags, and field information:

• Show tags for all measurements: show tag keys
• Show fields for all measurements: show field keys
• Show fields for all series: show series
• Show tags for measurement aic_cosnodestats: show tag keys from aic_cosnodestats
• Show fields for measurement aic_cosnodestats: show field keys from aic_cosnodestats
• Show series for measurement aic_cosnodestats: show series from aic_cosnodestats
• SHOW TAG VALUES WITH KEY = cosNodeID
• Show retention policies on stats_system

Use the following queries from the database shell to show COS node statistics:

• select * from aic_cosnodestats where time > now() - 1m
• select count(storageUsed) from aic_cosnodestats

Use the following HTP queries from the database shell against InfluxdB:

• https://172.20.235.82:8443/db/query?&u=admin&p=default&db=stats_system&pretty=true&q=show%20measurements
• https://172.20.235.82:8443/db/query?&u=admin&p=default&db=stats_system&pretty=true&q=select+*+from+aic_cosnodestats+where+cosNodeID=%27335574603%27
• http://172.22.120.80:8086/query?&u=admin&p=default&db=stats_system&pretty=true&q=select * from aic_cosnodestats where cosNodeID='335574603' and time > now() - 6d

Alarms at Master Nodes

Alarms at Worker Nodes

Statistics at Master and Worker Nodes