====================================================================================================
SYSBENCH BENCHMARK REPORT 
====================================================================================================
MACHINE:  drizzle-build-n01.wc1.dfw1.stabletransit.com
RUN ID:   713
RUN DATE: 2010-12-05T01:36:54.929536
WORKLOAD: innodb_1000K_readonly
SERVER:   drizzled
VERSION:  staging
REVISION: 1975
COMMENT:  1975: Lee Bieber 2010-12-04 [merge] Merge Padraig - Small fix needed 
when building a plugin out-of-tree.
====================================================================================================

TRENDING OVER LAST 5 runs 
Conc   TPS     % Diff from Avg   Diff       Min        Max        Avg        
STD       
====================================================================================================
16     1786.37       -1.04%     -18.77    1768.50    1833.59    1805.14      
18.02
32     2404.17       +2.44%      57.36    2225.70    2434.25    2346.81      
54.82
64     2395.40       -0.45%     -10.82    2315.99    2482.15    2406.22      
53.64
128    2435.35       +0.79%      19.09    2386.91    2466.70    2416.26      
25.85
256    2437.35       +0.73%      17.56    2388.36    2484.44    2419.79      
32.91
512    2212.91       +0.15%       3.28    2193.78    2231.28    2209.63      
11.16
1024   1303.74       -3.07%     -41.33    1291.35    1404.22    1345.07      
34.37
====================================================================================================

TRENDING OVER Last 20 runs 

Conc   TPS     % Diff from Avg   Diff       Min        Max        Avg        
STD       
====================================================================================================
16     1786.37       -0.03%      -0.55    1724.98    1833.59    1786.92      
30.28
32     2404.17       +1.59%      37.65    2225.70    2468.90    2366.52      
55.27
64     2395.40       -0.75%     -18.07    2287.26    2536.57    2413.47      
54.22
128    2435.35       -0.23%      -5.53    2386.91    2511.48    2440.87      
29.04
256    2437.35       +0.23%       5.68    2388.36    2484.44    2431.67      
23.73
512    2212.91       +0.21%       4.59    2179.54    2249.04    2208.32      
15.20
1024   1303.74       -2.36%     -31.44    1264.57    1447.72    1335.18      
37.08
====================================================================================================

TRENDING OVER ALL runs 

Conc   TPS     % Diff from Avg   Diff       Min        Max        Avg        
STD       
====================================================================================================
16     1786.37      -11.93%    -241.95    1724.98    2233.42    2028.33     
104.00
32     2404.17       -9.60%    -255.42    2010.47    2909.92    2659.59     
136.91
64     2395.40      -11.45%    -309.79    2191.91    2998.80    2705.19     
158.24
128    2435.35       -8.93%    -238.90    2107.33    2911.98    2674.24     
115.40
256    2437.35       -3.79%     -96.13    2174.91    2691.02    2533.48      
85.86
512    2212.91      +17.28%     326.04     968.50    2470.12    1886.87     
503.47
1024   1303.74      +13.46%     154.67     485.25    1866.15    1149.07     
490.92
====================================================================================================
FULL REVISION COMMENTARY:

  1974.1.2: Lee Bieber 2010-12-04 [merge] Merge Padraig - Small fix needed when 
building a plugin out-of-tree.
    1971.3.4: Padraig O'Sullivan 2010-12-04 Put the schema and table header 
files in the correct directive...
    1971.3.3: Padraig O'Sullivan 2010-12-03 Actually correct the right file 
this time...
    1971.3.2: Padraig O'Sullivan 2010-12-03 Correct indentation.
    1971.3.1: Padraig O'Sullivan 2010-12-03 Make sure that schema and table 
headers in message directory are installed since storage_engine header includes 
them.
  1974.1.1: Lee Bieber 2010-12-04 [merge] Merge Joe - More work on storing 
transaction log in innodb
    1802.17.27: Joseph Daly 2010-12-03 [merge] merge trunk
    1802.17.26: Joseph Daly 2010-12-03 update POTFILES.in
    1802.17.25: Joseph Daly 2010-12-02 remove replication test, this 
functionality is tested in the transaction_log tests
    1802.17.24: Joseph Daly 2010-12-02 fix non deterministic test
    1802.17.23: Joseph Daly 2010-12-01 test fixes
    1802.17.22: Joseph Daly 2010-12-01 clean up diffs
    1802.17.21: Joseph Daly 2010-12-01 revert change
    1802.17.20: Joseph Daly 2010-12-01 better error handling in 
transaction_reader
    1802.17.19: Joseph Daly 2010-12-01 sort in asc order
    1802.17.18: Joseph Daly 2010-12-01 [merge] merge trunk
    1802.17.17: Joseph Daly 2010-12-01 port additional tests
    1802.17.16: Joseph Daly 2010-11-30 move more tests to use inno replication 
log
    1802.17.15: Joseph Daly 2010-11-30 insert/alter tests and added new option 
to transaction_reader
    1802.17.14: Joseph Daly 2010-11-29 fix a problem where a create table A, 
and insert into A were having a trx number that was lower for the insert then 
the create.
    1802.17.13: Joseph Daly 2010-11-27 disable by default
    1802.17.12: Joseph Daly 2010-11-27 fix up transaction reader to work with 
inno replication log
    1802.17.11: Joseph Daly 2010-11-18 use trx_id from message
    1802.17.10: Joseph Daly 2010-11-17 [merge] merge trunk
    1802.17.9: Joseph Daly 2010-11-11 enable innodb replication by default for 
testing
    1802.17.8: Joseph Daly 2010-11-11 fix up transaction log column
    1802.17.7: Joseph Daly 2010-11-10 [merge] merge steward innodb fixes
      1802.19.1: Brian Aker 2010-11-10 [merge] Merge Stewart
        1802.18.1: Stewart Smith 2010-11-10 migrate innobase 
insert_replication_mesasge to instead use the INNOBASE API directly instead of 
through the InnoDB parser. This means it should be reentrant. We construct our 
own prebuilt structure, set the row data and write the row as part of the 
transaction (as before). We currently do *NOT* handle any row lock conflicts. 
i.e. if you ever write code that can block the insert into the table, this will 
explode horribly. You also probably want to take note that the create table 
statement for the innobase parser only creates 4 byte integers, so i've had to 
just take the low 32bits of txn id.
    1802.17.6: Joseph Daly 2010-11-09 only allocate a trx if one is not 
associated with the session
    1802.17.5: Joseph Daly 2010-11-09 fix up trans id code, still allocating 
multiple txns
    1802.17.4: Joseph Daly 2010-11-08 dont call trx_free_for_mysql() and call 
check_trx_exists() this will allocate if the trx does not exist
    1802.17.3: Brian Aker 2010-11-08 [merge] Merge in trunk.
    1802.17.2: Brian Aker 2010-10-01 Clean up some of the worst of the memory 
abuses in the plugin.
    1802.17.1: Brian Aker 2010-10-01 First pass over replicating to the guts of 
Innodb.

_______________________________________________
Mailing list: https://launchpad.net/~drizzle-benchmark
Post to     : drizzle-benchmark@lists.launchpad.net
Unsubscribe : https://launchpad.net/~drizzle-benchmark
More help   : https://help.launchpad.net/ListHelp

Reply via email to