ESP Trace Flags

Each ESP process may specify its own set of flags for controlling trace behavior. The specific flags may be shared with other platform components, shared amongst ESPs, or unique to an ESP.

Accepted Flags

Detailed description of flags used by any ESP.

Shared Platform Flags

Flags defined in system/jlib/jtrace.hpp and used by multiple platform processes.

Flags will be added to this list as tracing logic is updated in ESP code. For example, the shared platform flag traceHttp is expected to be used, as are a number of ESP-specific options.

Shared ESP Flags

Flags defined in esp/platform/esptrace.h and applicable to most, if not all, ESP configurations.

traceDetail

Set the default trace level in the process. Accepted case-insensitive values are:

• 1, standard: most important output
• 2, detailed: average verbosity output
• 3, max: highest verbosity output
• default: use existing level
  • Release builds default to standard
  • Debug builds default to max
• 0, none, all other values: no trace output

Process Configuration

Containerized

esp.yaml

Each ESP application's configuration object may embed one instance of a traceFlags object. Within this object, at most one property per accepted flag is expected. Properties not described here are ignored.

For example, the eclwatch process might be configured to use detailed reporting like this:

esp:
- name: eclwatch
  traceFlags:
    traceDetail: 2

Cluster Overrides

A values file may be used with the helm install command to override the settings of all ESPs. The --set option may be used to target the settings of a specific ESP in the configured array.

Bare-Metal

No support for defining trace flags is provided by the configmgr application. Within a stand-alone esp.xml file, however, a traceFlags child of the EspProcess element may be defined.

The previous YAML example may be reproduced in XML with the following:

<EspProcess ...>
  <traceFlags traceDetail="2" />
  ...
<EspProcess>