[infra-disco] branch main updated: Add a README.md

[Ian Young]

This is an automated email from the git hooks/post-receive script.

iay pushed a commit to branch main
in repository infra-disco.

View the commit online:
http://git.shibboleth.net/view/?p=infra-disco.git;a=commit;h=239e49422e6f3a92da092fc7cf3654462a31417a

The following commit(s) were added to refs/heads/main by this push:
       new  239e494   Add a README.md
239e494 is described below

commit 239e49422e6f3a92da092fc7cf3654462a31417a
Author: Ian Young <ian at iay.org.uk>
AuthorDate: Wed Sep 9 11:56:18 2020 +0100

    Add a README.md
---
 README.md | 40 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 40 insertions(+)

diff --git a/README.md b/README.md
new file mode 100644
index 0000000..1a655b9
--- /dev/null
+++ b/README.md
@@ -0,0 +1,40 @@
+# `infra-disco`
+
+This project contains the discovery feed generation process used by the
+`shibboleth.net` infrastructure.
+
+The repository is deployed in a private user context called `disco`.
+The script `refresh` is run once every four hours using this `crontab`
+entry:
+
+```
+17 */6 * * * infra-disco/refresh
+```
+
+The main operation in the script is running the `DiscoFeed` pipeline included
+in the `discoshibb.xml` configuration file:
+
+```bash
+$MDA_CMD --quiet $LOCATION/discoshibb.xml DiscoFeed
+```
+
+If successful, the MDA invocation creates a local output file called
+`discofeed.json`.
+
+The final output from the script, however, is a similarly named file in the
+parent directory. The final output file is served by Apache.
+
+The final output file is only overwritten if (a) the MDA run is successful
+and (b) the new local output file is different than the final output file.
+This is intended to improve the caching behaviour of browser clients.
+
+The `discoshibb.xml` configuration combined five sources of metadata:
+
+* Local metadata describing the `shibboleth.net` entities,
+* Local metadata describing the ProtectNetwork identity provide,
+* External metadata from three national federations.
+
+Each of the external metadata sources are fetched in a separate processing
+pipeline; each aggregate document is validated against a known signing key for
+the individual source. The three external sources are then merged together
+with duplicates being discarded.

-- 
To stop receiving notification emails like this one, please contact
the administrator of this repository.