Overview
Check-in intervals determine how long it takes for a policy or power-state changes at the server to reach the affected devices. This article describes the intervals and illustrates the steps involved in these changes.
Environment
Surveyor 6
Information
Time Lags Occurring Before Changes Appear in the Administrator Console
When you change devices' power state in the Administrator Console, the change might not be immediately reflected in the console. This can occur if you perform either of the following options:
- Change the power state manually on selected devices.
- Modify policy settings.
- Change devices' Wake on WAN's proxy preferences.
The reason is that check-in (polling) intervals are set on server components so that devices check for updates at a configured interval. This is done to reduce the amount of network traffic and server load. Some of these intervals are built-in and can be modified.
Determining and Modifying Check-In Intervals
Settings that affect the time required for a power state or policy change to take effect are configured at the server level. In some cases, you might want to be aware of the maximum amount of time a policy change might take and plan changes based this time in mind.
On the other hand, Device check-in interval is a server setting that you can modify in the Administrator Console. By default, the Device check-in interval is 10 minutes, and actual check-in times are jittered among all devices, i.e., each time a device checks in, it is assigned a next check-in time between 0.5 and 1.5 times the check-in interval, and designated times are uniformly distributed among devices. This avoids large numbers of devices having the same check-in times, overloading the server each time they need to check in.
Important: Be aware that every time a device checks in, a 128-byte line is generated in the Internet Information Service (IIS) log file. The lower the check-in interval time, the higher the amount of log data that will be created daily. To reduce the amount of data being logged, set the check-in interval to a higher number of minutes.
To modify the interval, on the Configure menu , click System Settings and change the value as you see fit (see the example below).
Verdiem.Enterprise.Common.config File Settings
This .config file contains the following check-in interval settings:
Setting |
Description |
EnergywiseProxyCheckinIntervalSeconds setting |
Determines how frequently the EnergyWise Proxy Server polls the Surveyor server for pending commands to forward to EnergyWise devices. |
WolProxyCheckinIntervalSeconds |
Affects the time for a Wake on WAN proxy to forward a wake request to target computers. |
You can find the Verdiem.Enterprise.Common.config file in the C:\Program Files\Verdiem\Surveyor\Configuration path. The default value for both of these settings is 120 seconds. They are formatted as follows:
<add key="EnergywiseProxyCheckinIntervalSeconds" value="120"/>
Note: If you change the values, you will need to uncomment the settings.
Details: How a Policy Update Reaches Devices
The below image explains what happens when you make policy changes, such as editing a power scheme or assigning a new policy. Numbers in the diagram represent the corresponding steps in the table.
Step |
Description |
Time required |
1. |
The Surveyor agent checks in with the Surveyor server for policy or power state updates. |
0.5–1.5 times the value set for the Device check-in interval in the Administrator Console. |
2. |
The agent obtains and applies any updates. |
Immediate (seconds) |
3. |
The agent sends policy/status updates to the server. |
Immediate (seconds) |
4. |
The server updates the database and the user interface (UI) elements. |
Immediate (seconds) |
Details: How a Manual Wake Request Reaches Computers
This section assumes that the wake request is sent through Wake on WAN to the agent computers. Any of the following methods can send the manual wake request:
- Right-clicking selected devices in the Administrator Console and choosing the Wake command.
- Using Wake for Remote Access to specify the name of the computer to wake.
- From a Microsoft Configuration Manager console, using the Verdiem Power Management Pack.
NOTE: Although it does not directly affect the time required to wake computers manually, Wake for Remote Access has its timeout setting for determining when a computer is not reachable.
Step |
Description |
Time required |
1. |
The server receives the wake request. |
Immediate (seconds) |
2. |
The server notifies the Wake on WAN proxy of the wake request, sending the target computers' Media Access Control (MAC) addresses to the proxy. |
Immediate (seconds) |
3. |
The Wake on WAN proxy sends a magic packet to the subnets of the target computers. |
Immediate (seconds) |
4. |
The target computer wakes from the low-power state (including off) upon receiving the magic packet. |
Immediate (seconds) |
5. |
The Surveyor agent starts and checks in with the server (following the normal check-in process). |
Immediate (seconds) |
6. |
The server updates the interface from which the wake command originated. |
Immediate (seconds) |
You can change the Wake on WAN proxy check-in interval in the Verdiem.Enterprise.Common.config file as described in Determining and Modifying Check-In Intervals section of this article.
Details: Manually Transitioning Computers to a Low-Power State
This section describes the process for a direct shutdown or sleep command sent to agents from the Administrator Console.
Step |
Description |
Time required |
1. |
The Surveyor client agent checks in with the server for policy changes. |
0.5–1.5 times the value set for the Device check-in interval in the Administrator Console. |
2. |
The agent carries out the low-power state transition instructions from the server. The server is not explicitly notified that the computer has transitioned to low-power. |
The power-state transition occurs upon receipt of instructions. |
3. |
After 1.5 times the Device check-in interval, the server instructs the Administrator UI to change the agent status to Not checking in. |
Up to 1.5 times the value set for |
Details: Policy Update or Power-State Change on EnergyWise Devices
This section describes the process for agent-less EnergyWise devices, where agent-less means that the Surveyor client agent is not installed on the device, but the device is part of an EnergyWise domain.
Step |
Description |
Time required |
1. |
The change is made through the Administrator Console. |
Immediate (seconds) |
2. |
The EnergyWise Proxy Server polls the Surveyor server and looks for pending commands. |
1–2 minutes.1 |
3. |
The EnergyWise Proxy Server retrieves and queues any commands to be processed. |
Immediate (seconds) |
4. |
The EnergyWise Proxy Server sends queued commands to the designated primary switch for the EnergyWise domains and waits for them to return their device status. |
Immediate pending load |
5. |
The EnergyWise primary switch sends the IOS command show energywise to all other switches in the EnergyWise domains for interface and device status. |
Immediate (seconds) |
6 |
Each EnergyWise switch receives the command from the primary switch and then polls its interfaces for status. |
The time required depends on the number of switches, interfaces, and devices in the EnergyWise domain. |
7. |
The EnergyWise switches send polling results back to the primary switch. |
Immediate (seconds) |
8. |
The primary switch passes the status it receives from the other switches to the EnergyWise proxy server. |
Immediate (seconds) |
9. |
The proxy server passes the switch, interface, and device status to the Surveyor server. |
Immediate (seconds) |
10. |
The server updates the database and UI to reflect the status of the devices. |
Immediate (seconds) |
This is determined by the EnergywiseProxyCheckinIntervalSeconds setting described in the Determining and Modifying Check-In Intervals.
The provisioning server web.config file also contains a timeout value set to 10000 milliseconds. The proxy server waits this amount of time for the switches to return their device status. The timeout is reset each time a switch reports back. If the timeout value is reached, the proxy server ignores any subsequent communication resulting from the original request.
If the timeout value (in the provisioning server web.config file) has been exceeded at the time the proxy server receives the results, the information will be ignored, and the device status will not be updated in the Administrator Console until the next check-in interval.
Comments
0 comments
Article is closed for comments.