-
ENABLE_CORE_DUMP
: Controls whether a core dump file is created in the event of a crash. Allowed values are true and false. The default is false. -
ROX_ENABLE_AFTERGLOW
: Allows to enable afterglow functionality to reduce networking usage. See the corresponding Afterglow section for more details. The default is true. -
ROX_NETWORK_DROP_IGNORED
: Ignore connections with configured protocol and port pairs (at the moment onlyudp/9
). The default is true. -
ROX_IGNORE_NETWORKS
: A coma-separated list of network prefixes to ignore. Any connection with a remote peer matching this list will not be reported. The default is169.254.0.0/16,fe80::/10
-
ROX_NON_AGGREGATED_NETWORKS
: A coma-separated list of network prefixes indicating endpoints which should never be considered for aggregation. This option can be useful when the CIDR blocks used for services or PODs are not standard private subnets, as it will prevent Collector from handling them as public IPs. -
ROX_NETWORK_GRAPH_PORTS
: Controls whether to retrieve TCP listening sockets, while reading connection information from procfs. The default is true. -
ROX_COLLECTOR_DISABLE_NETWORK_FLOWS
: Allows to disable processing of network system call events and reading of connection information from procfs. Mainly used in case of network-related performance degradation. The default is false. -
ROX_PROCESSES_LISTENING_ON_PORT
: Instructs Collector to add information about the originator process on all network listening-endpoint objects. The default is false. -
ROX_COLLECT_CONNECTION_STATUS
: Instruct Collector to track the network connections status. With this enabled, advertising of asynchronous connections will be postponed until their status is known and they are successful. The default is true. -
ROX_COLLECTOR_ENABLE_CONNECTION_STATS
: Instructs Collector to harvest and publish metrics regarding the network connections handled by the connection tracker. The data is summarized in quantiles published by prometheus. This is enabled by default.-
ROX_COLLECTOR_CONNECTION_STATS_QUANTILES
: a coma separated list of decimals defining the quantiles for all connection metrics. Default:0.5,0.90,0.95
-
ROX_COLLECTOR_CONNECTION_STATS_ERROR
: the allowed error for the quantiles (used to aggregate observations). Default:0.01
-
ROX_COLLECTOR_CONNECTION_STATS_WINDOW
: the length of the sliding time window in minutes. Default:60
-
-
ROX_COLLECTOR_SINSP_CPU_PER_BUFFER
: Allows to control how many sinsp buffers are going to be allocated. The resulting number of buffers will be calculated as the overall number of CPU cores available divided by this value. The default value is 1, meaning one buffer for each CPU. The value 0 is special, it instructs sinsp to allocate only one buffer no matter how many CPUs are there. This parameter affects CO-RE BPF only. -
ROX_COLLECTOR_SINSP_BUFFER_SIZE
: Specifies the size of a sinsp buffer in bytes. The default value is 8MB. This value must be a power of 2, a multiple of the system page size and greater than2 * page_size
. -
ROX_COLLECTOR_SINSP_TOTAL_BUFFER_SIZE
: Specifies the allowed total size of all sinsp buffer in bytes. If the actual value will be larger than that due to number of available CPUs,ROX_COLLECTOR_SINSP_BUFFER_SIZE
will be adjusted to match the limit. The default value is 512 MB and based on the default memory limit specified for Collector DaemonSet in ACS. -
ROX_COLLECTOR_SINSP_THREAD_CACHE_SIZE
: Puts upper limit on how many thread info objects are going to be kept in memory. Since for process-based workloads it's the main part of memory consumption, this value effectively translates into the upper limit for memory usage. Note, that Falco puts it's own upper limit on top of that, which is 2^17. -
ROX_COLLECTOR_ENABLE_DETAILED_METRICS
: Specified whether to expose per-syscall metrics. This information could be useful for troubleshooting, but under normal functioning is quite verbose. The default is true. -
ROX_COLLECTOR_INTROSPECTION_ENABLE
: Enable the introspection API and publish the corresponding endpoints. With this API, it is possible to dump some of the internal state of Collector. Refer to the troubleshooting section for more details. The default is false. -
ROX_COLLECTOR_EXTERNAL_IPS_ENABLE
: Manually enables or disables the external IPs feature. When enabled, Collector will send full details about IPs communicating with the cluster. When disabled, external IPs are aggregated as an "External sources" bucket (unless they match a defined CIDR-block). Runtime configuration is the preferred method to control external IPs, and it overrides this variable. Default is disabled.
NOTE: Using environment variables is a preferred way of configuring Collector, so if you're adding a new configuration knob, keep this in mind.
Collector configuration could be passed via --collector-config
argument to
the binary and is represented as JSON string.
-
scrapeInterval
: Specifies how frequently network scraping is performed, in seconds. The default value is 30 seconds. -
turnOffScrape
: Turn off network scraping. The default is false. -
logLevel
: Sets logging level. The default is INFO.
The external IPs feature can be enabled or disabled using a file or ConfigMap. This is an optional method and does not have to be used. This file overwites the ENABLE_EXTERNAL_IPS feature flag. When using collector by itself a file can be mounted to it at /etc/stackrox/runtime_config.yaml. The following is an example of the contents
networking:
externalIps:
enable: true
perContainerRateLimit: 1234
Alternatively, if collector is used as a part of Stackrox, the configuration can be set using a ConfigMap. The following is an example of such a ConfigMap.
apiVersion: v1
kind: ConfigMap
metadata:
name: collector-config
namespace: stackrox
data:
runtime_config.yaml: |
networking:
externalIps:
enable: true
perContainerRateLimit: 1234
The file path can be set using the ROX_COLLECTOR_CONFIG_PATH
environment variable.
Whenever the configuration file is updated or created, collector will update
the configuration. If the configuration file is deleted the configuration will revert
to the default. The configuration file can be created after collector start up.
-
--collection-method
: Which technology to use for data gathering. Either "ebpf" or "core-bpf". -
--grpc-server
: GRPC server endpoint for Collector to communicate, in the form "host:port".