Multiple collectors can help monitor isolated network segments and provide redundancy if a collector becomes unavailable.
We recommend giving each collector a clear description so you can easily identify its location and purpose.
When to deploy multiple collectors
You may want to deploy multiple collectors when:
- You have multiple physical sites that cannot reach one another.
- Network segments are physically or logically isolated.
- You want a backup collector if the primary collector’s host becomes unavailable.
- You need to maintain monitoring during collector, host, or network-path outages.
Multiple collectors can support failover when each collector can communicate with the subnets and devices it may need to monitor.
How multiple collectors support failover
If multiple collectors are configured for the same network, Auvik can use another eligible collector when the assigned collector goes offline.
If one collector loses power, connectivity, or host availability:
- A standby collector may take over monitoring.
- Device status updates and device-level alerts can continue if the standby collector can reach the affected devices.
- When the primary collector returns and is healthy, Auvik can reassign the network back to the primary collector.
A standby collector does not help if it cannot reach the monitored subnet or devices.
For example, if the collector assigned to subnet A goes offline, a collector assigned only to an isolated subnet B will not take over monitoring for subnet A.
How to manage collector assignments
You can manage collector assignments from:
Discovery > Manage Networks
When adding or editing a network, you can choose automatic or manual collector selection.
Automatic collector selection
With automatic selection, Auvik chooses an eligible collector that can communicate with the subnet.
If the assigned collector goes offline, another eligible collector may take over, provided that it can reach the subnet.
Automatic selection can cause monitoring gaps or false alerts if a collector is selected that cannot access the subnet. Confirm that all eligible collectors have the required network reachability before using automatic selection.
Manual collector selection
With manual selection, you can define:
- A primary collector
- One or more standby collectors
- The priority of each standby collector
To configure a primary and standby collector:
- Go to Discovery > Manage Networks.
- Add a new network or edit an existing network.
- Set Auvik Collector Selection to Manual.
- Select the Primary Collector.
- Select Add Standby Collector(s).
- Select one or more standby collectors.
- Set the priority for each standby collector.
- Click Save.
Auvik uses the primary collector whenever it is online and healthy. If the primary collector goes offline, Auvik fails over to a configured standby collector. When the primary collector returns and is healthy, Auvik reassigns the network back to the primary collector.
You can configure a tertiary standby collector by adding another standby collector and assigning its priority.
Collector reachability requirements
Every primary and standby collector assigned to a network must be able to reach the subnets and devices it may need to monitor.
Confirm that the appropriate access is available for the required monitoring protocols, such as:
- SNMP
- SSH
- Telnet
- WMI
- API access
- ICMP
The collector must also be able to communicate with the Auvik cloud.
If a standby collector cannot reach a device, it cannot provide monitoring or alerting for that device during a primary collector outage.
When multiple collectors help maintain alerting
Multiple collectors can help preserve device-level alerting when:
- The primary collector loses power or connectivity.
- A collector host or virtual machine becomes unavailable.
- One network path to a device fails but another collector has a working path.
- A site has separate network segments that require independent monitoring.
- A standby collector has the required credentials and protocol access.
For example, if Collector A is monitoring a device and becomes unavailable, Collector B can continue polling the device if Collector B can reach the device and is configured as an eligible standby collector.
When multiple collectors do not help
Multiple collectors will not maintain device-level alerts when:
- All collectors that can reach the device are offline.
- The standby collector cannot reach the device’s subnet.
- Routing, ACL, VPN, or firewall changes block the standby collector.
- The device is reachable only through the failed collector’s network path.
- The collectors are assigned to different isolated subnets and are not configured to provide coverage for one another.
During an outage affecting all eligible collectors, Auvik may stop receiving current device status information. Device-level alerts may be delayed or unavailable until a collector resumes monitoring.
Collector alerts and device-level alerts
The collector disconnected alert indicates that a collector has lost communication with the Auvik cloud and may no longer be monitoring its assigned networks.
A collector disconnected alert does not guarantee that every device-level alert will stop or continue. Device-level alerting depends on whether:
- Another eligible collector is configured.
- The alternate collector is online.
- The alternate collector can reach the affected devices.
- The required monitoring credentials and protocols are available.
- The network path between the alternate collector and the devices is working.
Configure collector health notifications so you know when monitoring redundancy has been reduced.
Network traffic considerations
A primary and standby collector configuration is designed to provide failover without unnecessary concurrent polling.
To reduce unnecessary management traffic:
- Use one primary collector for each monitored network.
- Configure additional collectors as standby collectors when redundancy is required.
- Avoid assigning multiple active collectors to the same network unless parallel monitoring is intentional.
- Segment networks between collectors when separate monitoring paths are required.
- Confirm that standby collectors have sufficient resources to handle the additional workload during failover.
If multiple collectors are actively polling the same devices, the environment may generate additional SNMP, SSH, WMI, API, or ICMP traffic.
Collector placement best practices
Place primary and standby collectors in separate failure domains whenever possible.
Examples include:
- Separate power circuits
- Separate physical hosts
- Separate hypervisors
- Separate network paths
- Separate data centers or sites
Avoid placing both collectors on the same host, power source, or network path if that dependency could fail at the same time.
Give each collector a description that identifies its location and purpose. Examples include:
- Calgary primary collector
- Calgary standby collector
- Head office collector
- Disaster recovery collector
Testing collector failover
Test failover during a planned maintenance window using a non-production device or network where possible.
- Confirm that the primary and standby collectors are online.
- Confirm that both collectors can reach the test network and device.
- Temporarily stop the primary collector or block its path to the test network.
- Confirm that the standby collector assumes responsibility.
- Verify that monitoring continues.
- Confirm that expected device-level alerts continue to generate.
- Restore the primary collector or network path.
- Confirm that Auvik returns monitoring to the primary collector after it is healthy.
Only perform this test if you are authorized to stop the collector or modify the network path. Coordinate with your network administrator before testing.
Troubleshooting missing alerts during a collector outage
If device-level alerts do not fire as expected:
- Confirm which collector was assigned to the affected network.
- Check whether the primary collector was offline or unable to reach the devices.
- Confirm that a standby collector was configured.
- Verify that the standby collector was online.
- Confirm that the standby collector could reach the affected subnets and devices.
- Review routing, ACL, VPN, and firewall access between the standby collector and the devices.
- Confirm that required monitoring credentials and protocols are available.
- Review collector resource usage and polling errors.
- Check whether all eligible collectors were unreachable during the outage.
Network, firewall, VPN, routing, ACL, and device-access changes should be reviewed by an authorized network or device administrator.
Adding an additional collector
To add an additional collector:
- Go to Auvik Collectors.
- Click Add Auvik Collector.
- Select the appropriate deployment method.
- Deploy the collector on a host with access to the required networks.
- Wait for the collector to initialize.
- Approve the collector in Auvik.
- Confirm that the collector appears as connected.
- Assign the collector to the appropriate networks.
- Configure it as a primary or standby collector as required.
New collectors must be approved before they can communicate with the Auvik cloud and begin monitoring.
Replacing a collector
When replacing a collector:
- Deploy the replacement collector.
- Approve the replacement collector.
- Assign it to the required network or networks.
- Confirm that discovery and monitoring are working.
- Power down the old collector.
- Reject or delete the old collector if it is permanently retired.
Do not delete a collector that may reconnect. A deleted collector that is still powered on and connected can reappear in Auvik and require approval.