Clarify all/staged mirror types

Signed-off-by: Joerg Jaspert <joerg@debian.org>

1 files changed, 6 insertions, 1 deletions

diff --git a/etc/runmirrors.mirror.sample b/etc/runmirrors.mirror.sample
index 744ad98..3df490d 100644
--- a/etc/runmirrors.mirror.sample
+++ b/etc/runmirrors.mirror.sample
@@ -11,8 +11,13 @@
 #
 # Type is either all, staged or mhop, meaning:
 # all    - do a "normal" push. Trigger them, go on.
+#          This is the default for a Debian mirror. It does all the
+#          right things (when the mirror uses ftpsync), like staged-syncing.
 # staged - do a two-stage push, waiting for them after stage 2(and all others that
-#          are staged) before doing stage2
+#          are staged) before doing stage2.
+#          Contrary to all, this waits for ALL downstream mirrors
+#          marked "staged", so most useful if they are behind a DNS RR
+#          and need to update dists/ at near the same time.
 # mhop   - send a multi-hop staged push. This will tell the mirror to initiate
 #          a mhop/stage1 push to its staged/mhop mirrors and then exit.
 #          When all mhop got back we then send the stage2 through to them.