Monitoring Cloud Databases#
Monitoring is available for Cloud Databases to allow users to get more information about their database instances via the monitoring checks that are available. The monitoring capability now applies to all instances, not just those that were created after the v1.0.27 release on July 29, 2013.
There is a 1:1 mapping of Cloud Databases to monitoring entities. The
monitoring entity has a field called agent_id
that is the same as the
Cloud Database instance id
.
A user can list the monitoring checks available on the monitoring entity by using the List Checks for an entity operation available in the Rackspace Monitoring API.
The following monitoring checks are available to users:
Load Average
CPU
Memory
Storage (Disk)
Network
Database
The following data points are supported on the Database checks:
core.aborted_clients
core.queries
core.connections
core.uptime
threads.created
threads.running
threads.connected
qcache.queries_in_cache
qcache.lowmem_prunes
qcache.free_memory
qcache.total_blocks
qcache.inserts
qcache.hits
qcache.not_cached
qcache.free_blocks
innodb.buffer_pool_pages_total
innodb.buffer_pool_pages_flushed
innodb.row_lock_time_avg
innodb.rows_inserted
innodb.row_lock_time_max
innodb.rows_updated
innodb.buffer_pool_pages_free
innodb.buffer_pool_pages_dirty
innodb.row_lock_time
innodb.rows_deleted
innodb.rows_read
A user can exercise these checks by submitting a Test an existing check API request using the Rackspace Monitoring service.
Monitoring alarms available to users include:
Storage alarm sent via email to the default admin notification list
label = Low Disk Space
WARNING > 90% used
message emits = Less than 10% free space left.
CRITICAL > 95% used
message emits = Less than 5% free space left.
Note
Users are not allowed to modify the managed entity or checks.
Users are allowed to add notifications and alarms to the existing checks.
For email notifications, users should check their junk/spam folder in case they have filtering on their email messages.
Refer to the following information on the monitoring API for more details: