How do I configure my system to transmit mini-SEED data using SEEDlink?

Scope

This document applies to systems running the Platinum operating system. This includes acquisition systems such as

EAMs and NAMs;
digitisers such as Affinity systems and DM24SxEAMs;
and
digital instruments with embedded acquisition modules such as 3TDEs, 40TDEs and 5TDEs.

Terminology

The SEEDlink protocol requires that the data-consumer contacts the data-provider and requests data. This is known as a "pull" connection because the consumer initiates a connection down which it then pulls data. The data-provider does not "push" data to data-consumers unless they request it.

In this document, we will refer to the data-provider as the server, transmitter or sender and to the data-consumer as the client or receiver.

Overview

We first configure the mini-SEED compressor (gdi2miniseed), which packetises the data from the central multiplexer (gdi-base) and we then create a SEEDlink server instance (seedlink-out) with automatic channel naming. Lastly, we filter or map channel names, if required.

Because SEEDlink can be used to talk to many different systems, we're going to use the IRIS slinktool as the test receiver in these examples. This is the reference implementation for the protocol so, if slinktool works but your system doesn't, the fault is probably with your system.

You can download slinktool for Linux, free of charge, from the IRIS website at
ds.iris.edu/ds/nodes/dmc/software/downloads/slinktool/

Let's get started

Configuring the compressor

In the web interface, visit Home → Configuration → Data handling → Services → Mini-SEED → gdi2miniseed

The resulting screen shows a list of all Mini-SEED compressor instances that have been configured:

Although the default instance is marked as does not start automatically, it will be started if a service that requires it is started. We will configure a SEEDlink server later in this procedure: that will require the compressor, so we don't need to configure the compressor to start automatically.

The SEEDlink protocol requires mini-SEED blocks to be 512 Bytes long. If you need blocks of a different size for any reason, such as local recording, you will need a dedicated compressor; if not, you can use the default compressor. Therefore:

If you do not want to record data locally in mini-SEED format, click on Mini-SEED compressor. Default instance - does not start automatically.
If you do want to record data locally in mini-SEED format and you are happy for your recorded data to have 512-Byte blocks, click on Mini-SEED compressor. Default instance - does not start automatically.
If you do want to record data locally in mini-SEED format and you want the recorded data to have a block size which is not 512 Bytes, click on Create new service instance

On the General tab, check that the block-size is set to 512 Bytes. This is essential: SEEDlink does not work with other block-sizes.

If you are already recording Mini-SEED data with 512-Byte block-sizes and you have already configured channel name mappings, you can skip the next step. Otherwise, click over to the Channels tab and set Naming mode to Automatic:

(We can revisit Channel name mapping later, if required.)

Scroll to the bottom of the page and click if you have made any changes.

Configuring a SEEDlink server instance

Next, in the web interface, visit Home → Configuration → Data handling → Services → seedlink-out

If you have no instances configured, click on Create new service instance; Otherwise, click on the instance that you have created.

Make sure that the Enable check-box is ticked. The Server hostname/IP address field should be left blank. The Server port/service name should be set to seedlink unless you have a reason to specify a different listener port.

If you created a new compressor instance earlier (so that you could record with a different block size), you should select your new instance in the Mini-SEED converter drop-down menu.

Scroll to the bottom of the page and click to save your changes.

Next, visit Home → Control → Services and scroll to the bottom of the page.

Click for the seedlink-out.0 service.

Next, visit Home → Summary → System status.

The tab for Mini-SEED Compressor. Default instance should have a green indicator (). It will have started automatically when you started the SEEDlink server. The tab for SEEDlink network server (instance 1) should have an amber indicator () because it has no clients at the moment.

Mapping Channel Names

The system will provide a default mapping between the GCF channel names and the mini-SEED channel names. This is normally adequate but, if you want to use different names for the mini-SEED channels, you can change the mapping. This is easy to do if you run the system in automatic name mapping mode for a few minutes; this allows the system to pre-populate the table with a list of the incoming channels. This is why we have left channel mapping until now.

If you don't want to use the default mapping for channel names or if you don't want to transmit all channels, you should configure this now. To do this, in the web interface, visit Home → Configuration → Data handling → Services → Mini-SEED → gdi2miniseed

The system will display the list of configured compressors. Select the one that you configured earlier and click over to the Channels tab.

The Naming mode drop-down menu offers three choices:

Automatic - all channels are compressed and named automatically. This mode offers no filtering and provides system-generated names for each channel.
Semi-automatic - all channels are compressed, names may be mapped below. In this mode, one or more of the channels may be renamed by editing entries in the mapping table.
Manual - only channels named below are compressed. This mode offers both channel filtering and name mapping.

Clicking

on any row will open a new row. In the same way, rows can be deleted by clicking the corresponding

button.

If the form is submitted when the table is full, extra blank lines are appended.

When you have finished making changes, scroll to the bottom of the page and click to save your map.

Testing

The testing described here uses slinktool, which is the reference implementation for a SEEDlink client. We have configured the SEEDlink transmitter to use the default SEEDlink TCP port (18000) so we do not need to specify that during the tests.

Station list

The -L argument to slinktool requests a station list from the server and prints it. You need to specify the IP address or the DNS name of your Platinum system in place of 10.10.0.2 below. Depending on your configuration, you should get one or two lines of output, like this:

user@example:~$ slinktool -L 10.10.0.2 A1 VEL0 Station A1 ACC0 Station user@example:~$

If the name(s) of the station(s) is not what you expected, revisit the name mapping described above and ensure that you clicked to save any changes.

Packet dump

The -p argument to slinktool prints header details of each packet received. You need to specify the IP address or the DNS name of your Platinum system in place of 10.10.0.2 below.

user@example:~$ slinktool -p 10.10.0.2 A1_VEL0_02_CHN, 412 samples, 500 Hz, 2017,073,11:13:55.000000 (latency ~31.0 sec) A1_VEL0_02_CHN, 412 samples, 500 Hz, 2017,073,11:13:55.824000 (latency ~30.2 sec) A1_VEL0_02_CHN, 412 samples, 500 Hz, 2017,073,11:13:56.648000 (latency ~29.3 sec) A1_VEL0_02_CHN, 412 samples, 500 Hz, 2017,073,11:13:57.472000 (latency ~28.5 sec) A1_VEL0_02_CHN, 412 samples, 500 Hz, 2017,073,11:13:58.296000 (latency ~27.7 sec) A1_VEL0_02_CHN, 412 samples, 500 Hz, 2017,073,11:13:59.120000 (latency ~26.9 sec) A1_VEL0_02_CHN, 412 samples, 500 Hz, 2017,073,11:13:59.944000 (latency ~26.1 sec) … user@example:~$

Type ctrl + C to stop the printing.

If both of these tests are successful, the configuration of your transmitter is complete and you can start to configure your receiver.

Once your client has connected successfully, its details are displayed in the SEEDlink network server tab of the main status display. If multiple clients connect to your server, they are listed as shown below:

If you want different clients to see different selections of streams or different mapped names, you will need to configure multiple server instances, each listening on a different port.

If you have any problems or questions, please contact for help and advice.