RealHeavyDude
Well-Known Member
OpenEdge 11.3.1 Solaris SPARC 64 Bit
I am having trouble understanding this. Maybe I am completely wrong - but here goes:
I have a database, blocksize is 8192, for which I moved the Schema Area to the B2. The Schema Area only contains the meta schema of the database and consist of one variable extent with a size of 18 MB on disk.
Therefore I thought that some 10'000 blocks should be way more than necessary. But here I am wrong
If I have 3.4 OS million reads - I would need 3.4 million buffers - rly??!@#$%%% - 27 GB for the schema area?
Now I am really puzzled. Okay - I run a dbanalys - but ...
Can anybody point me into a direction to understand this?
Thanks in Advance,
RealHeavyDude.
I am having trouble understanding this. Maybe I am completely wrong - but here goes:
I have a database, blocksize is 8192, for which I moved the Schema Area to the B2. The Schema Area only contains the meta schema of the database and consist of one variable extent with a size of 18 MB on disk.
Code:
proenv>proutil d_onedb -C viewB2 | grep "Buffer Pool"
Area "Schema Area":6 - Alternate Buffer Pool
Area "LOB 01":34 - Primary Buffer Pool
Area "Index_01":100 - Primary Buffer Pool
Area "Index_02":110 - Primary Buffer Pool
Area "Index_03":120 - Primary Buffer Pool
Area "Index_04":130 - Primary Buffer Pool
Area "Index_05":140 - Primary Buffer Pool
Area "Index_06":150 - Primary Buffer Pool
Area "Index_07":160 - Primary Buffer Pool
Area "Index_08":170 - Primary Buffer Pool
Area "Data_01":200 - Primary Buffer Pool
Area "Data_02":210 - Primary Buffer Pool
Area "Data_03":220 - Primary Buffer Pool
Area "Data_04":230 - Primary Buffer Pool
Area "Data_05":240 - Primary Buffer Pool
Area "Data_06":250 - Primary Buffer Pool
Area "Data_07":260 - Primary Buffer Pool
Area "Data_08":270 - Primary Buffer Pool
svkeady:t302218$ ls -lh
total 301552574
-rw-r--r-- 1 uug001 gug001 18M Nov 17 16:00 d_onedb.d1
-rw-r--r-- 1 uug001 gug001 2.0G Nov 17 12:11 d_onedb_100.d1
-rw-r--r-- 1 uug001 gug001 2.0G Nov 17 12:11 d_onedb_100.d2
-rw-r--r-- 1 uug001 gug001 2.0G Nov 17 12:11 d_onedb_100.d3
Therefore I thought that some 10'000 blocks should be way more than necessary. But here I am wrong
Code:
Alternate Buffer Pool
Logical reads 669121K 2566539 42775.65 1930085.52
Logical writes 5 0 0.00 0.01
O/S reads 3406599 12760 212.67 9596.05
O/S writes 7 0 0.00 0.02
Marked to checkpoint 0 0 0.00 0.00
Flushed at checkpoint 0 0 0.00 0.00
Writes deferred 2 0 0.00 0.01
LRU2 skips 0 0 0.00 0.00
LRU2 writes 0 0 0.00 0.00
APW enqueues 0 0 0.00 0.00
Alternate buffer pool hit ratio: 99 %
LRU2 replacement policy enabled.
If I have 3.4 OS million reads - I would need 3.4 million buffers - rly??!@#$%%% - 27 GB for the schema area?
Now I am really puzzled. Okay - I run a dbanalys - but ...
Can anybody point me into a direction to understand this?
Thanks in Advance,
RealHeavyDude.