Cluster Status - Field Help
Configuration Settings
Edit any of the values on the Cluster Status panel
and click Update Configuration.
- Default sticky time
- The default sticky time for all future ports on this cluster.
Sticky time is used to create an affinity relationship between a client and a specific server.
After the sticky time has been exceeded, the client may be sent to a server different from the previous.
Sticky time may be overridden on a per-port basis. The default value is 0, meaning that the
port is not sticky. This can also be set at the executor level.
To enable SSL ID affinity, you must set the default sickytime to a non-zero number and then add a new port
with content-based routing forwarding method protocol. To disable SSL ID affinity on the port, set the
stickytime for the port to zero.
- Default port stale timeout
- The default number of seconds during which there can be no activity on a connection before that
connection is removed. It may be overridden on a per-port basis. The default for port 21 is 900; the
default for port 23 is 32,000,000. The default value for all other ports is 300 and can also be set at
the executor level.
- Default port weight bound
- The default weight bound for all future ports on this cluster. It may be overridden on a per-port
basis. The default value is 20 and can also be set at the executor level.
- Maximum number of ports
- The maximum number of ports on this cluster. The default value is 8 and can also be set at the
executor level.
- Default maximum servers per port
- The default maximum number of servers per port on this cluster. It may be overridden on a per-port
basis. The default value is 32 and can also be set at the executor level.
- Default port protocol
- The default port type for all future ports on this cluster. It may be overridden on a per-port basis.
Possible values are TCP, UDP, or both protocols (TCP/UDP) and can also be set at
the executor level. The default value is
TCP/UDP.
- Shared bandwidth
- The level of bandwidth (kilobytes per second) to be shared at the cluster or executor level.
Sharing bandwidth at the cluster level allows a port (or ports)
to share a maximum amount of bandwidth across several ports (applications/protocols) within the same
cluster. Sharing bandwidth at the executor level allows
a cluster (or clusters) within the entire Dispatcher configuration to share a maximum amount of
bandwidth. The default value is 0. If the value is zero, then
bandwidth cannot be shared.
Note: You should specify a maximum shared bandwidth value that does not exceed the total
bandwidth (total server capacity) available.
- Primary host for the cluster
- The nonforwarding address of the primary host for the cluster in dotted decimal format.
Current Statistics
- Cluster address
- The dotted decimal IP address of this cluster.
- Cluster alias on interface
- Specifies whether or not the cluster has been aliased. Possible results are: configured, unconfigured, or unavailable.
- SYNs on active connections
- The number of redundant connection requests. A high rate of growth in this statistic could indicate slow or no response from servers in this cluster.
- SYNs on finished connections
- The number of requests for new connections, where those new connections are the same as current connections in FIN state.
- FINs received before SYNs
- FINs dropped due to no connection. Rapid increase in this statistic could indicate that the stale timeout is too low. Any percentage greater than 1% of total packets forwarded could indicate a problem.
- ACKs received before SYNs
- ACKs dropped due to no connection. Rapid increase in this variable could indicate that the stale timeout is too low. Any percentage greater than 1% of total packets forwarded could indicate a problem.
- RSTs received before SYNs
- RSTs dropped due to no connection. Rapid increase in this variable could indicate that the stale timeout for this cluster is too low. Any percentage greater than 1% of total packets forwarded could indicate a problem.
- Total packets dropped
- Packets dropped due to no connection. Rapid increase in this variable could indicate that the stale timeout for the ports on this cluster is too low. Any percentage greater than 1% of total packets forwarded could indicate a problem.
- SYN on SYN for quiesced server
- The number of requests for new connections, where those new connections are the same as connections already received for a server that has been quiesced.
- KBytes transferred per second
- The number of kilobytes transferred in a one second interval.
- Connections per second
- The current number of connections per second for this cluster.
Proportions
- Proportion given to active connections:
- A number from 0-100 representing the proportion of weight to be given to the active connections.
The default value is 50. The weight of the active connections, plus the weight of new connections,
plus the weight of the advisor information, plus the weight of the system monitoring program
information must total 100.
- Proportion given to new connections:
- A number from 0-100 representing the proportion of weight to be given to the new connections. The
default value is 50. The weight of the active connections, plus the weight of new connections, plus
the weight of the advisor information, plus the weight of the system monitoring program information
must total 100.
- Proportion given specific to the port:
- A number from 0-100 representing the proportion of weight to be given to the information coming
from an advisor about a port. The default value is 0. The weight of the active connections, plus the
weight of new connections, plus the weight of the advisor information, plus the weight of the system
monitoring program information must total 100.
- Proportion give to system metrics:
- A number from 0-100 representing the proportion of weight to be given to the information from the
system metrics, such as from metric server. The default value is 0. The weight of the active connections, plus
the weight of new connections, plus the weight of the advisor information, plus the weight of the
system monitoring program information must total 100.
- Sum of proportions:
- The sum of the above proportions. This total must equal 100.
- List of system metrics
- Current list of metric values and their proportions
To remove a system metric, highlight the system metric in the list and click remove.
Lists
- List of Ports
- Current list of ports, the number of servers running on those ports and the port type: TCP, UDP or
TCP/UDP which is chosen at the time the port is added to the server.
To remove a port, highlight the port in the list and click remove.
To refresh the list at any time, click refresh.
|