If there is only a single pool available for restage,
the selection engine can end up in a loop not being
able to select this pool. This has been reported first
by IN2P3 for 1.7. The problem is understood but the
bug is obviously still in 1.8. NDGF is reporting it
as well now.
-- original request --------------------
Here we go again. I just did a
06.26.41 000100000000000003380758 : Adding Object :
[Ljava.lang.Object;@dd34aa
06.26.41 000100000000000003380758 : Starting Engine
06.26.41 000100000000000003380758 : ACTIVATING STATE ENGINE
000100000000000003380758 43803409
06.26.41 000100000000000003380758 : StageEngine called in mode
Suspended with object retry
06.26.41 000100000000000003380758 : StageEngine left with : Init
(Continue)
06.26.41 000100000000000003380758 : StageEngine called in mode Init
with object (NULL)
06.26.41 PFL [000100000000000003380758] : calculateFileAvailableMatrix
_expectedFromPnfs : []
06.26.41 queryPoolsForPnfsId : number of valid replies : 0
06.26.41 PFL [000100000000000003380758] : calculateFileAvailableMatrix
_acknowledgedPnfsPools : []
06.26.41 PFL [000100000000000003380758] : calculateFileAvailableMatrix
: db matrix[*,0] [pool-phenix-reco-sn014a]
06.26.41 PFL [000100000000000003380758] : calculateFileAvailableMatrix
: av matrix[*,0] []
06.26.41 000100000000000003380758 : askIfAvailable : file not in pool
at all
06.26.41 000100000000000003380758 : askIfAvailable : Took 43803456
06.26.41 000100000000000003380758 : StageEngine left with : Staging
(Continue)
06.26.41 000100000000000003380758 : StageEngine called in mode Staging
with object (NULL)
06.26.41 queryPoolsForCost : costModule : pool-phenix-reco-sn014a
(1384216436)
pool-phenix-reco-sn014a={Tag={{hostname=ccdcacsn014}};size=1384216436;SC=0.0018688492964084105;CC=0.0;}
06.26.41 000100000000000003380758 : askFor cache : Second shot
excluding : pool-phenix-reco-sn014a
06.26.41 000100000000000003380758 : CacheException(rc=150;msg=No cheap
candidates available for 'cache')
06.26.41 000100000000000003380758 : Selection cache took : 43803456
06.26.42 000100000000000003380758 : StageEngine left with : Init (Wait)
That's the only thing I have in the pinboard
