White Papers

Dell HPC Lustre Storage solution with Mellanox Infiniband EDR
The max_pages_per_rpc parameter is a tunable that sets the maximum number of pages that
will undergo I/O in a single RPC to that OST.
[root@node057 ~]# lctl set_param osc.*.max_rpcs_in_flight=64
The max_rpcs_in_flight is a tunable that sets the maximum number of concurrent RPCs in
flight to the OST. This parameter in majority of cases will help with small file IO patterns.
[root@node057 ~]# lctl set_param llite.*.max_read_ahead_mb=1024
The max_read_ahead_mb is a tunable that sets the maximum amount of data readahead on a
file. These are read ahead in an RPC sized chunk.
[root@node057 ~]# lctl set_param osc.*.max_dirty_mb=1024
The max_dirty_mb is a tunable that sets how much dirty cache per OST can be written and
queued. This generally benefits large memory IO workloads.
[root@node057 ~]# lctl set_param osc.*.checksums=0
This tunable will disable checksums and overhead associated.
[root@node057 ~]# lctl set_param llite.*.max_cached_mb=40960
The max_cached_mb tunable is the maximum amount of inactive data that will be cached by
the client.
[root@node057 ~]# lctl set_param llite.*.max_read_ahead_per_file_mb=1024
The max_read_ahead_per_file_mb is a tunable that sets maximum read ahead for each file.
[root@node057 ~]# lctl set_param debug=0
This tuning parameter will disable debugging completely and free the associated overhead.
[root@node057 ~]# lctl set_param subsystem_debug=0
The subsystem_debug parameter will control the debugging logs3 for subsystems.
5. Conclusions
There is a well-known requirement for scalable, high-performance clustered file system solutions. The
Dell HPC Lustre Storage with Intel EE for Lustre addresses this need with a well-designed solution that
is easy to manage and fully supported. The solution includes the added benefit of the Dell PowerEdge™
13
th
generation server platform, PowerVault™ storage products and Lustre
®
technology, the leading
open source solution for a parallel file system. The Intel Manager for Lustre (IML) unifies the
management of the Lustre File System and solution components into a single control and monitoring
panel for ease of use.
The scale of the raw storage, 960TB per Object Storage Server Pair and up to 15GB/s of write and
18GB/s of read throughput in a packaged component, is consistent with the needs of the high
performance computing environments. The Dell HPC Lustre Storage solution is also capable of scaling in
throughput as easily as it scales in capacity.