Skip to main content

API client libraries and command line tools for the ESnet Monitoring Daemon (esmond).

Project description

Client programs for perfSONAR data


A discovery tool to quickly see what tests have been stored in an esmond perfSONAR archive. Give a list of tests in MA with the following sample information:

tool_name: bwctl/tracepath,traceroute
event_type: packet-trace, failures, path-mtu


Similar to get-endpoints, but this will fetch the actual metadata test data from an esmond perfSONAR archive. By default it will show the measurements that are common to all tests:


Including the –metadata-extended will also show the per-test measurements. This option can not be used with the CSV output option.

Sample default output:

tool_name: powstream

Sample output with the –metadata-extended flag:

tool_name: powstream
ip_transport_protocol: udp
sample_bucket_width: 0.0001
sample_size: 600
subject_type: point-to-point
time_duration: 60
time_interval: 0
time_probe_interval: 0.1


Tool to pull smaller, more focused sets of data from a perfSONAR MA. This requires a source/dest pair as well as a specific event type. Intended to be more of a “quick look” at some data. To gather more/larger amounts of data, esmond-ps-get-bulk is intended for that.


Tool to pull non-trivial amounts of data from a perfSONAR esmond archive.

Iterates through the metadata matching the user query and makes incremental data requests from the archive so as not to overwhelm the data store. When all of the data for a given event type associated with a given metadata has been gathered, it will be written to disc in either json or csv format, with the format:


So one would end up with a set of output files that look like this:

After the file for the metadata/event-type has been written, it will continue to the next event-type or metadata as appropriate. The “human readable” output format is not available in this program.

While designed to not murder an MA with massive data queries, this command can return a lot of data, so it is recommended to limit the scope of your query by source, dest, event-type, etc.

General esmond-ps perfSONAR client usage

Core and/or required args

These args are common to all clients. See the –help flag to get a complete list of options.


Required on all programs. Just the base protocol://host:port is required. When querying a default perfSONAR install, it is not necessary to include the URI as well. For example given a MA access URL of:

It is only necessary to provide:

–src and –dest

Source and destination for the tests. Both are required for some of the clients. This is input as raw IP addresses.

–start-time and –end-time

If these args are not included, it will default to grabbing data from the previous 24 hours. Arg input is parsed by the Python dateutil library which will preform pretty intelligent guesses about incoming date formats. It will understand structured things like ISO datetime formats, and more organic ones like “January 1 2015” - if a time is not given, will default 00:00 am, etc.

See: To see the variety of date formats that it will accept.


Requires a valid measurement event type. The command line arg –list-events can be used to give a list of valid event types.

Sometimes required.

Additional filtering args

There are additional args that can be used to filter results as well:


These should be fairly self-explanatory.


An additional power user filter that takes the format:

--filter key:value

This will add filters to the query string that goes to the MA. This option can be used more than once to add multiple filters to the query string, invalid filters will be ignored.



Select the desired output format from the choices ‘human,’ ‘json’ and ‘csv.’ Default is human readable for viewing in a terminal. The human and csv options are not allowed in all circumstances.


Required by esmond-ps-get-bulk - specifies a directory to write output files to. Will default to the current working directory.


By default in the output, IP addresses (source, dest, agent, etc) will be converted to a human readable fully qualified domain name. Using the -ip flag will stop this conversion and display all hostnames as raw IP addresses.

Example perfSONAR command line client usage

esmond-ps-get-endpoints examples

Get a list of all tests over the last 24 hours available in a given MA, show src/dest as raw ip addresses:

esmond-ps-get-endpoints --url --ip

Find all the powstream test data in a given MA since the beginning of the year:

esmond-ps-get-endpoints --url --ip --start-time 'January 1' --tool powstream

esmond-ps-get-metadata examples

Show all test metadata for a given destination over the last 24 hours, displayed in CSV format:

esmond-ps-get-metadata --url --dest --output-format csv

Show more detailed metadata information from an MA for all bwctl/iperf3 tests involving a particular source since the beginning of the year, showing extended test metadata like test duration, interval, etc as a list of json objects:

esmond-ps-get-metadata --url --tool bwctl/iperf3 --src --metadata-extended --output-format json --start-time 'Jan 1'

esmond-ps-get examples

Retrieve the past 24 hours of packet trace data for a src/dest pair:

esmond-ps-get --url --src --dest --event-type packet-trace

Get throughput data starting at the beginning of the month (presuming the month is April) for a src/dest pair:

esmond-ps-get --url --src --dest --event-type throughput --start-time 'April 1'

esmond-ps-get-bulk examples

Pull all failures event-type information from an MA since the beginning of the year and write out to current working directory as a set of json files:

esmond-ps-get-bulk --url  --event-type failures --start-time 'January 1' --output-format json

Pull all data associated with a given source from the past 24 hours and write to a custom directory in CSV format:

esmond-ps-get-bulk --url  --src --output-format csv -D ~/Desktop/tmp

Pull data for all event types measured by the powstream tool since the start of March and write to a custom directory in json format:

esmond-ps-get-bulk --url  --tool powstream --start-time 'March 1' --output-format json -D ~/Desktop/tmp

Pull all the data in an MA for the past 24 hours and output to current working directory in json format:

esmond-ps-get-bulk --url --output-format json

Esmond perfSONAR data loading programs

There are also client programs for writing data to an MA. This requires that the user have write access to the esmond instance.

Core and/or required args

The following args are required/generally needed by all programs that write data to an MA.

–user and –key

Both of these args are required. It is the username and api key string that was generated on the MA to allow access to it.


The url of the MA. Format where http or https can be the prefix. Just host and port information, no uri information. Defaults to http://localhost:8080.


Used when the REST API has been deployed under Apache using a ScriptAlias directive/prefix. This would commonly be set to ‘esmond’ since the canned CentOS deployments use script alias of /esmond to allow other things to run on the webserver (ie: so the REST API is not the root of the webserver). The default value is ‘/’ - which will not perform any prefixing.


Utility to parse and load GridFTP data.

This will read the default gridftp logs, process the “Transfer stats” entries, and upload the results to the pS esmond backend as metadata and either throughput or failures event types. This has been expanded (using the –json flag) to read the new json formatted gridftp logs that contain additional event types like retransmits, iostat, etc.

The basic use case would that this script be run from cron periodically over the day to parse and load data from the gridftp logs into an esmond backend. The scanning code will write out the contents of the record that was last loaded as a python pickle file to disc. This state file is used to pick up from the point the last processing pass got to.

Basic usage: the following arguments are required for baseline operation:

esmond-ps-load-gridftp -f ~/Desktop/gridftp.log -U http://localhost:8000 -u mgoode -k api_key_for_mgoode

In addition to the flags outlined above, required args


The path to the logfile to process. The code will normalize the path, so relative paths are fine. No default.

Commonly used args


Specifies that the log indicate by the –file flag is the json-formatted GridFTP files.


The path to the pickle file the scanning code uses to store the “state” of the last record that has been processed. Code uses this to know where to pick up on subsequent scans. This defaults to ./load_grid_ftp.pickle or ./load_grid_ftp.json.pickle as appropriate - will probably want to change this to a fully qualified path somewhere.


Suppresses writing the pickle state file out when the file has been scanned. This would be used when manually/etc processing one or more log files where it is desired to just parse the contents of an entire static (ie: no longer being written to) file. Defaults to False - use this flag to suppress writing the state file.


Can be used to specify a directory to write a log from the program to. If this is not set (the default), then log output will go to stdout.

Optional content selection args

The gridftp logs contain information on the user, the file being sent and the volume being written to. Since these might be considered to be sensitive data, this information is not sent to the backend by default. The following flags can be set to send that information if desired:

-F (--file_attr): send gridftp-file/value of FILE
-N (--name_attr): send gridftp-user/value of USER (name)
-V (--volume_attr): send gridftp-volume/value of VOLUME

Other/development args


Will process a single value starting at the last record sent and stop. This is mostly used for development/testing to “step through” a file record by record. It will set the pickle state file to the single record sent before exiting.

Running from cron and dealing with rotated logs

When running from cron the script should be run with the required arguments enumerated above and set the –pickle arg to a fully qualified path, and the –file arg should point to the logfile. It can be run at whatever frequency the user desires as the code will pick up from the last record that was processed. When running from cron, the –log_dir arg should be set so the logging output is written to a file rather than sent to stdout.

Log rotation interfere with this if the code has not finished scanning a log before it is rotated and renamed. If the code is run on the “fresh” log, it will not find the last record that was processed. To deal with this, this script should also be kicked off using the “prerotate” hook that logrotated provides.

When running this as a prerotate job, the -D (–delete_state) flag should also be used. This will delete the pickle state file when the scan is done with the log before it is rotated. The state file is deleted so that when the next cron job runs on the new “fresh” log, it will just start scaning from the beginning and not try to search for a record that it won’t find.

Alternately if the user doesn’t need the data to be periodically loaded, one could opt to exclusively run this as a logrotated/prerotate job such that the entire log is processed in one throw before it is rotated. In that case the –dont_write flag should be used.


Utility to take json-formatted output from bwctl (–parsable flag) and load the data into an esmond MA.

Currently supported tool types:

  • iperf3


Primarily relies on the required command line args (–user, –key, etc) outlined above and piped input from the bwctl command:

bwctl -c -s -T iperf3 --parsable --verbose |& esmond-ps-pipe --user mgoode --key api_key_for_mgoode

The primary thing (other than using a -T <tool> that is supported) is that bwctl must be run with both the –parsable flag (which generates the json output) and also the –verbose flag. esmond-ps-pipe pulls important metadata from the –verbose output, and uses it to identify the json part of the output.

If the program is unable to extract the necessary metadata and a valid json payload from the piped input, it will log a fatal error and exit.

Shell redirection

Note the “|&” that redirects the output from bwctl to esmond-ps-pipe - both stdout and stderr need to be piped to esmond-ps-pipe. That should work on Csh and current versions of Bash. This may vary from shell to shell - for example, older versions of Bash might need to use “2>&1 |” or something similar. The short of it is, the shell-specific way to redirect both stdout and stderr from bwctl is necessary.

If an error that looks something like this is generated:

ts=2015-10-20 11:37:24,881 event=id_and_extract.error id=1445366244 could not extract tool_name
ts=2015-10-20 11:37:24,881 event=id_and_extract.error id=1445366244 could not extract input_source
ts=2015-10-20 11:37:24,881 event=id_and_extract.error id=1445366244 could not extract input_destination
ts=2015-10-20 11:37:24,881 event=main.fatal id=1445366244 could not extract metadata and valid json from input
ts=2015-10-20 11:37:24,882 event=main.fatal id=1445366244 exiting

It is likely that the redirection is not being executed properly because tool_name, input_source and input_destination are all read from the bwctl headers that are being written to stderr.

Optional args


Like esmond-ps-load-gridftp, this takes a –log_dir arg which specifies the directory that logging output should be written to. If not specified, logging output will got to stdout.

Event types


The following event types are extracted (as appropriate RE: TCP, UDP, streams, etc) from the iperf3 data:


API Client Libraries for perfSONAR data

The pS data can be queried, retrieved and posted to the esmond/cassandra backend via a REST interface. This is streamlined by the following libraries:


Initializing the query interface

The query libarary has two main “top level” classes: ApiFilters and ApiConnect. ApiFilters lets the user, through a series of properties, set the primary query criteria like time ranges, source, destination, etc. The following criteria properties can be set:

verbose (for debugging/extended output)

After the query criteria have been set in the ApiFilters object, that is passed to the ApiConnect object as one of the args.

The ApiConnect object takes the url of the REST interface as an argument, along with the filters object, and optional username and api_key arguments if the user is accessing restricted functionality of the REST interface (non-public data, getting around throttling restrictions, etc).

A complete example of setting this up:

from esmond.api.client.perfsonar.query import ApiConnect, ApiFilters

filters = ApiFilters()

filters.verbose = True
filters.time_start = time.time() - 3600
filters.time_end = time.time()
filters.source = ''
filters.tool_name = 'bwctl/iperf3'

conn = ApiConnect('http://localhost:8000/', filters)

NOTE: the default perfSONAR/esmond deployments use a WSGIScriptAlias of /esmond prefixing the URI - this is set in Apache. The client libraries default to using this. But if one is doing development against the django runserver dev server, or if this has been set up differently, then the optional kwarg “script_alias” will need to be set as well. Against the dev server, it can be set to script_alias=None since the Apache directive is not in place.

Retrieving the data

The basic design of the returned data is a hierarchy of encapsulation objects that return additioanl objects objects, etc. All of the returned objects have informative __repr__ methods defined, that might help when doing initial development.

The top level call to the ApiConnect object is get_metadata(). This is an iterator that will return a series of Metadata objects matching the criteria given in the ApiFilters object. At the top level, the Metadata object exposes a series of properties giving additional information about the returned metadata. Example of this:

for md in conn.get_metadata():
    print md # debug info in __repr__
    print md.destination
    print md.ip_packet_interval

The following top-level properties are exposed by the Metadata object:

event_types (a list of event type names - more on this)

The next in the data object hierarchy is fetching the event types that are associated with the metadata. This can be done by either using an interator to access all of the event types:

for et in md.get_all_event_types():
    print et.event_type

or fetching a single one by name:

et = md.get_event_type('histogram-owdelay')

The top-level property “event_types” will return a list of valid event types that can be passed as the argument to get_event_type.

The EventType objects expose the following top-level properties:

summaries (a list of associated summaries - more on this)

The the actual underlying data are retrieved from the EventType objects by a call to the get_data() method, which returns a DataPayload object:

dpay = et.get_data()

The DataPayload object expose the following top-level properties:


The data_type property returns the underlying data_type in the payload, and the data property returns a list of DataPoint or DataHistogram objects as is appropriate. Both the DataPoint and DataHistogram objects expose the following properties:

ts (measurement timestamp as a UTC python datetime object)
val (the measurement or hisogram dict)
ts_epoch (the ts object expressed as UNIX time)

Putting it all together, to iterate throught all of the returned data:

for et in md.get_all_event_types():
    dpay = et.get_data()
    print dpay.data_type
    for dp in
        print dp.ts, dp.val

Some event types have aggregated summaries associated with them. Retrieving the summaries from an EventType object is very similar to pulling event types from a Metadata object. The following properties/methods are analogous to the ones that exist in the Metadata object:


This returns a list of two-element tuples: (summary-type, summary-window). The window is the time duration of the aggregation rollups.

The summary data can be retrieved by either using an iterator:

for summ in et.get_all_summaries():

Or a single type can be fetched:

summ = et.get_summary(summary-type, summary-window)

Like with the EventType object, the underlying data can be retrieved by calling get_data() to get a DataPayload object and call the data property on that to get a list of DataPoint objects.

Writing data to pS esmond/backend

The REST interface also supports adding metadata, event types and data if the user is properly authenticated using a username and api_key that has been generated by the admin of the system. The following are presented as an ordered process, but any single step of this can be done independently. The functionality for POSTing date can be found in the following libarary:

from import MetadataPost, \
    EventTypePost, EventTypeBulkPost

First one needs to create a new metadata entry - this is accomplished using the MetadataPost object. It is initialized with a REST url, username, api_key and a series of associated data - most required, a few optional (the commented key/val pairs in the arg dict are optional):

args = {
    "subject_type": "point-to-point",
    "source": "",
    "destination": "",
    "tool_name": "bwctl/iperf3",
    "measurement_agent": "",
    "input_source": "host1",
    "input_destination": "host2",
    # "time_duration": 30,
    # "ip_transport_protocol": "tcp"

mp = MetadataPost('http://localhost:8000/', username='pS_user',
    api_key='api-key-generated-by-auth-database', **args)

This will create the basic data associated with this metadata. Then add the event types and summaries associated with this metadata and post the new information:

mp.add_summary_type('packet-count-sent', 'aggregation', [3600, 86400])

new_meta = mp.post_metadata()

This writes the metadata information to the back end and returns the associated “read only” Metadata object that was covered in the previous section. This is mostly necessary to get the newly generated metadata_key property, it will be needed for other operations.

Next data can be added to the assocaited event types - the process is similar for both numeric and histogram data. Intialize an EventTypePost object similarly to the MetadataPost object, but also using the appropriate metadata_key and event_type to add the data to:

et = EventTypePost('http://localhost:8000/', username='pS_user',

Discrete data points can be added the process is similar for both numeric data and histogram data - first arg is an integer timestamp in seconds and the second is the value - and post it:

et.add_data_point(1397075053, 23)
et.add_data_point(1397075113, 55)

(or in the case of histograms)

et.add_data_point(1397075053, {28: 33})
et.add_data_point(1397075113, {9: 12})


It is also possible to bulk post data for a variety of event types associated with a single metadata using the EventTypeBulkPost interface. Intialize in a similar fashion minus the event_type arg:

etb = EventTypeBulkPost('http://localhost:8000/', username='pS_user',

Add a mix of data points specified by event type and post:

etb.add_data_point('time-error-estimates', 1397075053, 23)
etb.add_data_point('packet-loss-rate', 1397075053,
    {'numerator': 11, 'denominator': 33})

etb.add_data_point('time-error-estimates', 1397075113, 55)
etb.add_data_point('packet-loss-rate', 1397075113,
    {'numerator': 5, 'denominator': 8})


NOTE: as noted in the previous section, the optional script_alias kwarg works the same way with the POST interface.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

esmond_client-4.2.1.tar.gz (68.1 kB view hashes)

Uploaded Source

Built Distribution

esmond_client-4.2.1-py2-none-any.whl (69.2 kB view hashes)

Uploaded Python 2

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page