Update Edge Node Configuration Properties

Updated

Introduction

EVE-OS (Edge Virtualization Engine) provides a set of runtime configuration properties that allow you to customize edge node behavior from Edge Infrastructure Services without modifying the operating system itself. These properties control communication intervals, network behavior, security settings, storage allocation, logging verbosity, and system performance across edge nodes in your deployment.

Prerequisites

How It Works

Configuration properties are set through Edge Infrastructure Services and synchronized to the edge node during its next configuration check. The default check interval is 60 seconds but can be adjusted using timer.config.interval.

You can configure all the properties in ZCLI with the --config parameter, similar to the following example:

zcli edge-node update <EDGE_NODE> --config="timer.config.interval:30"

Most properties take effect immediately after the edge node receives the updated configuration. Some properties require an edge node reboot to take effect, as noted in the property descriptions.

Properties are applied as key-value pairs. Each property has a defined type, default value, and valid range. When a property has no minimum or maximum constraint, the corresponding column shows -.

Key Benefits

  • Control edge node communication frequency with Edge Infrastructure Services to optimize bandwidth usage across all managed nodes.
  • Enable or disable security features such as SSH access, VGA console output, and USB device support for targeted troubleshooting without system-wide exposure.
  • Adjust storage allocation between the base OS, edge containers, and ZFS to match the requirements of each deployment environment.
  • Configure logging verbosity and remote log upload to balance observability with bandwidth and storage costs.
  • Fine-tune network failover, download retry behavior, and cellular modem recovery for reliable edge operations in challenging connectivity environments.

Configuration Properties Reference

The following tables list all available EVE-OS runtime configuration properties, organized by category.

Application settings

Name Type Default Min Max Description
app.allow.vnc boolean false - - Allows access to EVE-OS VNC ports from external IP addresses. By default, only local access is allowed.
app.boot.order string "" - - Sets the device-wide default boot order for VMs. Supported values: "" (default UEFI behavior), "usb" (prioritize USB devices), "nousb" (remove USB devices from boot order). Can be overridden per VM.
app.fml.resolution string notset - - Sets the system-wide forced resolution for applications running in FML mode.
app.enable.tcp.mss.clamping boolean true - - Enables automatic TCP MSS clamping on forwarded edge container traffic to match the path MTU, preventing fragmentation on lower-MTU links.
process.cloud-init.multipart boolean false - - Enables handling of MIME multi-part cloud-init data for VMs that do not process it themselves.

Communication intervals

Name Type Default Min Max Description
timer.config.interval integer in seconds 60 5 86400 Controls how frequently the edge node retrieves configuration from Edge Infrastructure Services. Requires a reboot to take effect.
timer.cert.interval integer in seconds 86400 60 4294967295 Controls how frequently the edge node checks for new Edge Infrastructure Services certificates.
timer.metric.interval integer in seconds 60 5 3600 Controls how frequently the edge node reports metrics to Edge Infrastructure Services.
timer.deviceinfo.interval integer in seconds 600 30 4294967295 Controls how frequently the edge node reports device information to Edge Infrastructure Services even when nothing has changed. Requires a reboot to take effect.
timer.metric.diskscan.interval integer in seconds 300 5 3600 Controls how frequently the edge node scans disks for metrics.
timer.hardwarehealth.interval integer in seconds 43200 21600 4294967295 Controls how frequently the edge node reports hardware health information (ECC, SMART) to Edge Infrastructure Services.
timer.hardwareinfo.interval integer in seconds 10800 10800 4294967295 Controls how frequently the edge node reports hardware information (SMART) to Edge Infrastructure Services. Deprecated.
timer.ntpsources.interval integer in seconds 600 60 4294967295 Controls how frequently the edge node reports information about NTP sources to Edge Infrastructure Services. Reports are also sent when the NTP peer list or peer fields change.

Network connectivity

Name Type Default Min Max Description
network.fallback.any.eth "enabled" or "disabled" disabled - - When set to "enabled", allows the edge node to attempt connectivity using any available Ethernet, WiFi, or LTE interface with DHCP if configured connections fail. Forcibly enabled during onboarding if no network configuration exists.
network.download.max.cost 0-255 0 0 255 Sets the maximum port cost allowed for downloads. Use this to prevent downloads over expensive connections such as LTE.
network.switch.enable.arpsnoop boolean true - - Enables ARP snooping on switch network instances to prevent ARP spoofing attacks.
network.local.legacy.mac.address boolean false - - Enables legacy MAC address generation for local network instances. Use only on edge nodes where changing MAC addresses would cause incorrect network configuration in edge containers.
timer.reboot.no.network integer in seconds 604800 120 4294967295 Sets the duration without cloud connectivity before the edge node automatically reboots.
timer.update.fallback.no.network integer in seconds 300 60 4294967295 Sets the duration without cloud connectivity before the edge node falls back to a previous configuration.
timer.port.georedo integer in seconds 3600 60 4294967295 Controls how frequently the edge node redoes IP geolocation.
timer.port.georetry integer in seconds 600 5 4294967295 Controls how frequently the edge node retries geolocation after a failure.
timer.port.testduration integer in seconds 30 10 3600 Sets the duration to wait for DHCP to assign an address during port testing.
timer.port.testinterval integer in seconds 300 300 3600 Controls how frequently the edge node retests the current port configuration.
timer.port.timeout integer in seconds 15 0 3600 Sets the timeout for each HTTP send during port testing.
timer.port.testbetterinterval integer in seconds 600 0 4294967295 Controls how frequently the edge node tests higher-priority port configurations to check availability.

Download and update settings

Name Type Default Min Max Description
timer.download.retry integer in seconds 600 60 4294967295 Sets the wait time before retrying a failed download.
timer.download.stalled integer in seconds 600 20 4294967295 Sets the duration after which a download with no progress is cancelled.
blob.download.max.retries integer 5 1 10 Sets the maximum number of download retries when image verification fails.
timer.send.timeout integer in seconds 120 0 3600 Sets the timeout for each HTTP send operation.
timer.dial.timeout integer in seconds 10 0 3600 Sets the maximum time allowed to establish a network connection.
timer.test.baseimage.update integer in seconds 600 30 3600 Sets the duration to test a new base OS image before committing to the update. If the edge node becomes unstable during this period, it automatically falls back to the previous base OS version.

Storage settings

Name Type Default Min Max Description
storage.dom0.disk.minusage.percent integer 20 20 80 Sets the minimum percentage of the persist partition reserved for the EVE-OS base system.
storage.zfs.reserved.percent integer 20 1 99 Sets the minimum percentage of the persist partition reserved for ZFS performance optimization.
storage.apps.ignore.disk.check boolean false - - When enabled, allows edge containers to create images larger than currently available disk space. Use with caution, as enabling this can lead to out-of-disk-space errors during edge container operation.
timer.gc.vdisk integer in seconds 3600 60 4294967295 Controls how frequently EVE-OS garbage collects unused edge container virtual disks.
timer.defer.content.delete integer in seconds 0 0 86400 When set to a non-zero value, keeps content trees available for reuse for the specified duration after deletion. Set to 0 to delete immediately.

Debug and troubleshooting settings

Name Type Default Min Max Description
debug.enable.usb boolean false - - Allows USB devices such as keyboards on the edge node.
debug.enable.vga boolean false - - Allows VGA console output on the edge node.
debug.enable.ssh authorized SSH key "" - - Allows SSH access to EVE-OS when an authorized SSH public key is provided. An empty string disables SSH access.
debug.enable.console boolean false - - Allows console access to EVE-OS. Requires a reboot to disable.
debug.enable.vnc.shim.vm boolean false - - Allows VNC access to the container application shim VM. Requires a reboot to disable.

Memory settings

Name Type Default Min Max Description
memory.apps.ignore.check boolean false - - When enabled, allows edge containers to use more memory than currently available. Use with caution, as enabling this can lead to out-of-memory conditions.
memory.vmm.limit.MiB integer 0 0 1073741824 Manually sets the overhead memory in MiB allocated for each running VMM.
memory-monitor.enabled boolean false - - Enables external memory monitoring and memory pressure event handling.
internal-memory-monitor.store.enabled boolean true - - Enables Internal Memory Monitor (IMM) data collection and CSV storage. When enabled, memory metrics are collected at regular intervals and stored at /persist/memory-monitor/output/memory_usage.csv.
internal-memory-monitor.analyze.enabled boolean true - - Enables IMM leak detection analysis using statistical methods to detect memory growth patterns. Requires internal-memory-monitor.store.enabled to be set to true.

Go runtime settings

Name Type Default Min Max Description
gogc.memory.limit.bytes integer 0 0 4294967295 Sets the Golang runtime soft memory limit in bytes.
gogc.percent integer 100 0 500 Sets the Golang runtime garbage collector target percentage.
gogc.forced.interval.seconds integer in seconds 10 0 1000 Sets the minimum interval for forced garbage collection. Set to 0 to disable forced GC.
gogc.forced.growth.memory.MiB integer in MiB 50 10 1024 Sets the minimum allocated memory growth in MiB required to trigger the next GC execution.
gogc.forced.growth.memory.percent integer 20 5 300 Sets the minimum allocated memory growth percentage from the last reclaim required to trigger the next GC execution.

Logging settings

Name Type Default Min Max Description
newlog.allow.fastupload boolean false - - Enables faster upload of gzip-compressed log files to Edge Infrastructure Services.
newlog.gzipfiles.ondisk.maxmegabytes integer in MiB 2048 10 4294967295 Sets the maximum disk space in MiB for storing compressed log files on the edge node.
log.dedup.window.size integer 0 0 4294967295 Sets the size of the log deduplicator sliding window in number of messages. Set to 0 to disable deduplication.
log.count.filenames string "" - - Comma-separated list of log filenames to count and log once instead of logging every occurrence. A single comma "," filters out all entries without a filename field.
log.filter.filenames string "" - - Comma-separated list of log filenames to filter out entirely. A single comma "," filters out all entries without a filename field.
vector.enabled boolean true - - Enables the Vector service for advanced log filtering and transformations.
vector.config string "" - - Full base64-encoded configuration file for Vector in YAML format. Modify only the transforms section.

Network diagnostics

Name Type Default Min Max Description
netdump.enable boolean true - - Enables publishing of network diagnostics as compressed archives to /persist/netdump on the edge node.
netdump.topic.preonboard.interval integer in seconds 3600 60 4294967295 Controls how frequently network diagnostics of the same topic can be published before the edge node is onboarded.
netdump.topic.postonboard.interval integer in seconds 86400 60 4294967295 Controls how frequently network diagnostics of the same topic can be published after the edge node is onboarded.
netdump.topic.maxcount integer 10 1 4294967295 Sets the maximum number of network diagnostics that can be published per topic. The oldest entry is removed when the limit is exceeded.
netdump.downloader.with.pcap boolean false - - Includes packet captures in network diagnostics for download requests. TCP segments carrying payload are excluded and the total PCAP size is limited to 64 MB.
netdump.downloader.http.with.fieldvalue boolean false - - Includes HTTP header field values in network diagnostics for download requests. Enable with caution, as this may expose secrets such as datastore credentials.
diag.probe.remote.http.endpoint string "http://www.google.com" - - Sets the remote HTTP endpoint used to assess network connectivity when Edge Infrastructure Services is not reachable. Used for diagnostics only. Set to an empty string to disable.
diag.probe.remote.https.endpoint string "https://www.google.com" - - Sets the remote HTTPS endpoint used to assess network connectivity when Edge Infrastructure Services is not reachable. Hostnames are required; IP addresses are not accepted. Set to an empty string to disable.

Location reporting

Name Type Default Min Max Description
timer.location.cloud.interval integer in seconds 3600 300 4294967295 Controls how frequently the edge node reports geographic location information to Edge Infrastructure Services.
timer.location.app.interval integer in seconds 20 5 3600 Controls how frequently the edge node reports geographic location information to edge containers via the local profile server and metadata server.

Edge container settings

Name Type Default Min Max Description
timer.boot.retry integer in seconds 600 10 4294967295 Sets the wait time before retrying a failed edge container boot.
timer.appcontainer.stats.interval integer in seconds 300 1 4294967295 Controls how frequently EVE-OS collects edge container statistics.
timer.vault.ready.cutoff integer in seconds 300 60 4294967295 Sets the maximum wait time for the vault to become accessible before triggering an edge node reboot.
force.fallback.counter integer 0 0 4294967295 Forces fallback to an alternative OS image when the counter value changes.

Operating modes

Name Type Default Min Max Description
maintenance.mode enabled or disabled none - - When set to "enabled", puts the edge node into maintenance mode and stops running edge containers.
airgap.mode enabled or disabled none - - When set to "enabled", configures the edge node to operate without connectivity to Edge Infrastructure Services and to be managed locally via the LOC (Local Operator Console).

Cellular modem settings

Name Type Default Min Max Description
wwan.query.visible.providers boolean false - - Enables periodic queries (once per hour) of visible cellular service providers. Results are published under WirelessStatus for every modem.
wwan.modem.recovery.watchdog boolean false - - Enables the watchdog for cellular modems. When enabled, the edge node automatically reboots if a modem firmware crash cannot be recovered.
wwan.modem.recovery.reload.drivers boolean false - - Enables automatic reload of MBIM/QMI/MHI drivers when a modem firmware crash occurs. This runs before the watchdog mechanism is triggered.
wwan.modem.recovery.restart.modemmanager boolean false - - Enables automatic ModemManager restart when a modem firmware crash occurs. This runs before the watchdog mechanism is triggered and can be combined with driver reload recovery.

Goroutine leak detection

Name Type Default Min Max Description
goroutine.leak.detection.threshold integer 5000 1 4294967295 Sets the goroutine count at which leak detection is triggered regardless of growth rate.
goroutine.leak.detection.check.interval.minutes integer in minutes 1 1 4294967295 Sets the interval in minutes between goroutine count measurements.
goroutine.leak.detection.check.window.minutes integer in minutes 10 10 4294967295 Sets the analysis window in minutes for leak detection. The window must contain at least 10 measurements (no less than 10 times goroutine.leak.detection.check.interval.minutes).
goroutine.leak.detection.keep.stats.hours integer in hours 24 1 4294967295 Sets the number of hours to retain goroutine stats for leak detection.
goroutine.leak.detection.cooldown.minutes integer in minutes 5 1 4294967295 Sets the cooldown period in minutes after leak detection is triggered. During this period, no stack traces are collected and only warning messages are logged.

Kubernetes settings

Name Type Default Min Max Description
kubernetes.drain.timeout integer in hours 24 1 4294967295 Sets the maximum number of hours allowed for Kubernetes to drain a node.

k3s.config.override


 

 string None    

Base64 encoded string of the contents of a k3s config.yaml.d file to override existing config or add config parameters to an EVE-k node.

To properly override existing config settings, the following rules must be followed:

Follow The Merge Rules with existing config: https://docs.k3s.io/installation/configuration#value-merge-behavior

Using K3s server config options: https://docs.k3s.io/cli/server

Using K3s agent config options: https://docs.k3s.io/cli/agent

Network authentication settings

Name Type Default Min Max Description
pnac.dhcp.reacquire.max.retries integer 4 0 8 Sets the maximum number of DHCP reacquire retries after a PNAC (802.1X) port authentication state change. EVE-OS retries with exponential backoff (2s, 4s, 8s, ...) until the IP subnet changes or the retry limit is reached. Set to 0 to disable DHCP reacquire.
dhcp.enable.vendorclassid boolean true - - Enables sending the DHCP Vendor Class Identifier (Option 60) to identify the device as EVE-OS. Disable this if your DHCP server rejects unknown vendor class IDs.
scep.retry.interval integer in seconds 300 60 3600 Sets the interval between retry attempts for certificates that failed to enroll or renew, or that returned PENDING from the SCEP server.

EdgeView settings

Name Type Default Min Max Description
edgeview.authen.publickey string "" - - Specifies SSH public keys for Edgeview client command authentication. The user must provide the path to the SSH private key in the client script. Separate multiple public keys with newline characters.

Prometheus metrics settings

Name Type Default Min Max Description
msrv.prometheus.metrics.rps integer 1 1 4294967295 Sets the maximum number of requests per second for the Prometheus metrics endpoint.
msrv.prometheus.metrics.burst integer 10 1 4294967295 Sets the maximum burst size for the Prometheus metrics endpoint.
msrv.prometheus.metrics.idletimeout.seconds integer in seconds 240 1 4294967295 Sets the idle timeout in seconds for the Prometheus metrics endpoint. The rate limit resets when the connection has been idle for this duration.

Log levels

EVE-OS controls log verbosity through separate settings for each component. Logs for EVE-OS microservices use logrus levels: panic, fatal, error, warning, info, debug, and trace. Logs for syslog and kernel components use syslog levels: emerg, alert, crit, err, warning, notice, info, and debug. All log levels accept none (disables logging) and all (enables all log levels).

All logs are saved locally in /persist/newlog/keepSentQueue/ and are subject to rotation based on the configured maximum total size.

Global log level settings

Name Type Default Min Max Description
debug.default.loglevel string debug - - Sets the default verbosity of logs produced locally by EVE-OS microservices. Can be overridden per microservice using agent.<agentname>.debug.loglevel.
debug.default.remote.loglevel string warning - - Sets the default verbosity of logs sent by EVE-OS microservices to Edge Infrastructure Services. Can be overridden per microservice.
debug.syslog.loglevel string info - - Sets the verbosity of syslog messages produced locally.
debug.syslog.remote.loglevel string info - - Sets the verbosity of syslog messages sent to Edge Infrastructure Services.
debug.kernel.loglevel string info - - Sets the verbosity of kernel log messages produced locally.
debug.kernel.remote.loglevel string info - - Sets the verbosity of kernel log messages sent to Edge Infrastructure Services.
debug.tui.loglevel string info - - Sets the log level for the EVE-OS Text UI (TUI) monitor. TUI logs are only available locally and are not sent to Edge Infrastructure Services. Valid values: OFF, ERROR, WARN, INFO, DEBUG, TRACE (case insensitive).

Per-microservice log level settings

You can override the global log level for individual EVE-OS microservices using the following property pattern.

Name Type Default Min Max Description
agent.<agentname>.debug.loglevel string - - - Overrides debug.default.loglevel for the specified microservice. The legacy format debug.<agentname>.loglevel is also supported.
agent.<agentname>.debug.remote.loglevel string - - - Overrides debug.default.remote.loglevel for the specified microservice. The legacy format debug.<agentname>.remote.loglevel is also supported.

The following microservices support per-microservice log level settings: newlogd, wwan, nodeagent, downloader, tpmmgr, client, vcomlink, executor, vaultmgr, baseosmgr, zedagent, verifier, wstunnelclient, zfsmanager, zedkube, ledmanager, faultinjection, zedmanager, nim, loguploader, watcher, volumemgr, zedrouter, msrv, domainmgr, diag, scepclient.

Next Steps

Was this article helpful?
0 out of 1 found this helpful