Mat*_*DBA 6 performance sql-server dbcc tempdb ssd
我正在一个新的 SSD 阵列上进行计时试验,该阵列同时运行 SQLIO 测试和 DB 还原和 DBCC CHECKDB 调用的实际工作负载。我发现我的 SQLIO 批处理生成的 IOPS 和吞吐量与我观察到的工作负载之间存在重大差异,工作负载仅请求我使用 SQLIO 能够观察到的一小部分,通常在 5,000 IOPS 范围内并产生不超过 400 MB/s 的吞吐量。
如果硬件有足够的容量来处理负载,那么 DBCC CHECKDB 将消耗多少资源事件是否存在固有限制?我可以尝试哪些设置来扩展 DBCC CHECKDB 对 CPU 和磁盘资源的使用?
以下是具体...
从 systeminfo
OS Name: Microsoft Windows Server 2012 R2 Standard
OS Version: 6.3.9600 N/A Build 9600
System Manufacturer: HP
System Model: ProLiant DL580 G7
System Type: x64-based PC
Processor(s): 4 Processor(s) Installed.
[01]: Intel64 Family 6 Model 46 Stepping 6 GenuineIntel ~1042 Mhz
Total Physical Memory: 131,062 MB
Network Card(s): 4 NIC(s) Installed.
[01]: HP NC375i Integrated Quad Port Multifunction Gigabit Server Adapter
SQL 服务器信息
Microsoft SQL Server 2008 R2 (SP2) - 10.50.4000.0 (X64) Jun 28 2012 08:36:30 Copyright (c) Microsoft Corporation Enterprise Evaluation Edition (64-bit) on Windows NT 6.2 (Build 9200: )
使用 SQLIO 的测试脚本,其中参数文件被定向到 3 TB XtremeIO 闪存阵列 LUN 上的 40 GB 测试文件
sqlio -kW -t8 -s120 -o8 -fsequential -b64 -BH -LS -Fparam.txt
sqlio -kR -t8 -s120 -o8 -fsequential -b64 -BH -LS -Fparam.txt
sqlio -kW -t8 -s120 -o8 -frandom -b8 -BH -LS -Fparam.txt
sqlio -kR -t8 -s120 -o8 -frandom -b8 -BH -LS -Fparam.txt
XtremeIO 阵列的规格
XtremIO - 1 Brick
Version: 2.2.3 build 25
Build id: 9585409:HEAD-release-2_2
SQLIO 运行的结果
C:\SQLIO>sqlio -kW -t8 -s120 -o8 -fsequential -b64 -BH -LS -Fparam.txt
sqlio v1.5.SG
using system counter for latency timings, 2211143 counts per second
parameter file used: param.txt
file L:\testfile.dat with 8 threads (0-7) using mask 0x0 (0)
8 threads writing for 120 secs to file L:\testfile.dat
using 64KB sequential IOs
enabling multiple I/Os per thread with 8 outstanding
buffering set to use hardware disk cache (but not file cache)
using specified size: 40000 MB for file: L:\testfile.dat
initialization done
CUMULATIVE DATA:
throughput metrics:
IOs/sec: 23118.54
MBs/sec: 1444.90
latency metrics:
Min_Latency(ms): 0
Avg_Latency(ms): 2
Max_Latency(ms): 9
histogram:
ms: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24+
%: 5 7 46 41 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
C:\SQLIO>sqlio -kR -t8 -s120 -o8 -fsequential -b64 -BH -LS -Fparam.txt
sqlio v1.5.SG
using system counter for latency timings, 2211143 counts per second
parameter file used: param.txt
file L:\testfile.dat with 8 threads (0-7) using mask 0x0 (0)
8 threads reading for 120 secs from file L:\testfile.dat
using 64KB sequential IOs
enabling multiple I/Os per thread with 8 outstanding
buffering set to use hardware disk cache (but not file cache)
using specified size: 40000 MB for file: L:\testfile.dat
initialization done
CUMULATIVE DATA:
throughput metrics:
IOs/sec: 25160.07
MBs/sec: 1572.50
latency metrics:
Min_Latency(ms): 0
Avg_Latency(ms): 2
Max_Latency(ms): 8
histogram:
ms: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24+
%: 24 33 12 7 7 9 6 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
C:\SQLIO>sqlio -kW -t8 -s120 -o8 -frandom -b8 -BH -LS -Fparam.txt
sqlio v1.5.SG
using system counter for latency timings, 2211143 counts per second
parameter file used: param.txt
file L:\testfile.dat with 8 threads (0-7) using mask 0x0 (0)
8 threads writing for 120 secs to file L:\testfile.dat
using 8KB random IOs
enabling multiple I/Os per thread with 8 outstanding
buffering set to use hardware disk cache (but not file cache)
using specified size: 40000 MB for file: L:\testfile.dat
initialization done
CUMULATIVE DATA:
throughput metrics:
IOs/sec: 153634.35
MBs/sec: 1200.26
latency metrics:
Min_Latency(ms): 0
Avg_Latency(ms): 0
Max_Latency(ms): 1
histogram:
ms: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24+
%: 100 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
C:\SQLIO>sqlio -kR -t8 -s120 -o8 -frandom -b8 -BH -LS -Fparam.txt
sqlio v1.5.SG
using system counter for latency timings, 2211143 counts per second
parameter file used: param.txt
file L:\testfile.dat with 8 threads (0-7) using mask 0x0 (0)
8 threads reading for 120 secs from file L:\testfile.dat
using 8KB random IOs
enabling multiple I/Os per thread with 8 outstanding
buffering set to use hardware disk cache (but not file cache)
using specified size: 40000 MB for file: L:\testfile.dat
initialization done
CUMULATIVE DATA:
throughput metrics:
IOs/sec: 181107.89
MBs/sec: 1414.90
latency metrics:
Min_Latency(ms): 0
Avg_Latency(ms): 0
Max_Latency(ms): 5
histogram:
ms: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24+
%: 100 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
小智 0
是的,在某些情况下,等待大部分是 100% OLEDB,并且硬件看起来空闲。就我而言,我尝试在具有空间索引的 26 GB 表上运行 DBCC CHECKTABLE。它会运行、运行、运行……我将它移至我的工作站(6 核 Zeon,16 GB,带 2 个 SDD),希望能完成它。它运行得更快,但是运行又运行....我尝试使用 SQL 2012、SP2、跟踪标志等。普通表上的 DBCC 完成速度比生产中快大约 7 倍,并且确实有效地使用了我的磁盘,所以我知道我的硬件有帮助。具有空间索引的表上的 DBCC 运行了一周多才失败。(我没有限制内存并让操作系统处于饥饿状态。我还有虚拟机和其他东西。)当它运行时,我的机器似乎几乎空闲。我无法识别瓶颈。它不是 CPU 或磁盘。我正在考虑为此提交一份错误报告。
也许您可以使用 DBCC CHECKTABLE 来查看是否有一组选择的表也具有此行为。
归档时间: |
|
查看次数: |
1138 次 |
最近记录: |