White Papers

Dell Storage for HPC with Intel Enterprise Edition 2.3 for Lustre sofware
16
3125
20
1000000
24
3125
14
1050000
32
3125
10
1000000
48
3125
7
1050000
64
3125
5
1000000
72
3125
5
1125000
96
3125
4
1200000
120
3125
3
1125000
128
3125
3
1200000
144
3125
3
1350000
168
3125
2
1050000
192
3125
2
1200000
216
3125
2
1350000
240
3125
2
1500000
Figure 15: File Metadata Operations
Figure 16 illustrates the file metadata results using MDtest.As shown in this graph, file create metadata
operations start with less than 547 OPS at 1 thread and scale to over 16K OPS with 240 concurrent
threads. This may be due to the Lustre locks needed on the MDT, but also those on the OSTs, since
using a stripe count of 24 had a significant decrease in performance. At 240 threads, we had 2
directories (-b 2) and 2.2 million files were created.
File stat metadata operations is overwhelmingly the lightest metadata operation of the three observed.
A single thread test yield over 2K OPS and scale to more than 500K OPS with 216 concurrent threads.
The increase in performance could be due to improvements made in Lustre version 2.5 metadata
operations. Also of note is the use of 15K RPM drives in the MDT volumes.
Removal of files is also limited by accesses to OSTs, similar to the create operation. However, the
remove operation has an advantage over the create operation when increasing in total threads, starting
with over 1.8K OPS at 1 thread and scaling to over100K OPS with increasing concurrent threads.
0
100000
200000
300000
400000
500000
600000
1 2 4 8 12 16 24 32 48 64 72 96 120 128 144 168 192 216 240
OPS
Threads
MDtest FIles Metadata — Dell Storage for HPC with Intel EE Lustre
Solution
File Create File Stat File Remove