aboutsummaryrefslogtreecommitdiffstats
path: root/tests
diff options
context:
space:
mode:
authorPau Espin Pedrol <pespin@sysmocom.de>2018-06-08 16:29:51 +0200
committerPau Espin Pedrol <pespin@sysmocom.de>2018-06-08 17:03:06 +0200
commit868c67fed34d3a5f0d9d1e5150073fee3d71c26c (patch)
treed09d12a1b025600ba4d1e85a8be801b9c2f58855 /tests
parenta1a2556c2ae56399a0d24ff19c521c2a182cd31c (diff)
compact AGCH queue: Drop too msg diff than IMM_ASS_REJ
We saw in a recent prod setup a BSC with saturated channels. Further investigation lead to a AGCH queue of 1000 (previous hard_limit) messages, most of them being regular IMM ASSIGN (non REJ). Hence, we also want to get rid of other messages in the AGCH queue (like regular IMM ASSIGN). Furthermore, In this scenario, sending IMM ASS REJ is as important as other messages given than nowadays we support dynamic wait indicatior (calculated based on chan load), which means if we reach the MS, we can tell it to wait for a long time to reach us again, which is desirable. Change-Id: I022b8948da8be13fb8f4bc36e7c9dab11c35fddb
Diffstat (limited to 'tests')
-rw-r--r--tests/agch/agch_test.ok2
1 files changed, 1 insertions, 1 deletions
diff --git a/tests/agch/agch_test.ok b/tests/agch/agch_test.ok
index 4e0a9e13..a506f451 100644
--- a/tests/agch/agch_test.ok
+++ b/tests/agch/agch_test.ok
@@ -19,5 +19,5 @@ T BCCH slots
50 28 14 28 28 28
Testing AGCH messages queue handling.
AGCH filled: count 720, imm.ass 80, imm.ass.rej 640 (refs 640), queue limit 32, occupied 101, dropped 0, merged 198, rejected 421, ag-res 0, non-res 0
-AGCH drained: multiframes 16, imm.ass 34, imm.ass.rej 12 (refs 42), queue limit 32, occupied 0, dropped 56, merged 198, rejected 421, ag-res 15, non-res 30
+AGCH drained: multiframes 4, imm.ass 2, imm.ass.rej 8 (refs 26), queue limit 32, occupied 0, dropped 92, merged 198, rejected 421, ag-res 3, non-res 6
Success