On Tue, Mar 17, 2015 at 06:29:49PM +0100, 'Klaus Aehlig' via ganeti-devel wrote:
These tests are based on the wrong assumption that
iteratively taking the most balanced position for the
next instance will result in a balanced cluster. See
the discussion on issue 509 for more details.
Signed-off-by: Klaus Aehlig <[email protected]>
---
test/hs/shelltests/htools-single-group.test | 11 -----------
1 file changed, 11 deletions(-)
diff --git a/test/hs/shelltests/htools-single-group.test
b/test/hs/shelltests/htools-single-group.test
index f8e629c..254989c 100644
--- a/test/hs/shelltests/htools-single-group.test
+++ b/test/hs/shelltests/htools-single-group.test
@@ -5,12 +5,6 @@
./test/hs/hbal -v -v -p --print-instances -t$T/simu-onegroup.standard
>>>= 0
-# hbal should not be able to balance
-./test/hs/hbal -t$T/simu-onegroup.standard
->>> /(Nothing to do, exiting|No solution found)/
->>>= 0
-
-
# tiered single-group tests
./test/hs/hinfo -v -v -p --print-instances -t$T/simu-onegroup.tiered
>>>= 0
@@ -18,11 +12,6 @@
./test/hs/hbal -v -v -p --print-instances -t$T/simu-onegroup.tiered
>>>= 0
-# hbal should not be able to balance
-./test/hs/hbal -t$T/simu-onegroup.tiered
->>> /(Nothing to do, exiting|No solution found)/
->>>= 0
-
# hcheck should not find reason to rebalance
./test/hs/hcheck -t$T/simu-onegroup.tiered --machine-readable
>>> /HCHECK_INIT_CLUSTER_NEED_REBALANCE=0/
--
2.2.0.rc0.207.ga3a616c
LGTM
