File Systems

Available file systems

On the DEEP-EST system, three different groups of file systems are available:

• the ​JSC GPFS file systems, provided via ​JUST and mounted on all JSC systems;

• the DEEP-EST (and SDV) parallel BeeGFS file systems, available on all the nodes of the DEEP-EST system;

• the file systems local to each node.

The users home folders are placed on the shared GPFS file systems. With the advent of the new user model at JSC (​JUMO), the shared file systems are structured as follows:

• $HOME: each JSC user has a folder under /p/home/jusers/, in which different home folders are available, one per system he/she has access to. These home folders have a low space quota and are reserved for configuration files, ssh keys, etc.

• $PROJECT: In JUMO, data and computational resources are assigned to projects: users can request access to a project and use the resources associated to it. As a consequence, each user can create folders within each of the projects he/she is part of (with either personal or permissions to share with other project members). For the DEEP project, the project folder is located under /p/project/cdeep/. Here is where the user should place data, and where the old files generated in the home folder before the JUMO transition can be found.

The DEEP-EST system doesn't mount the $SCRATCH and $ARCHIVE file systems from GPFS, as it is expected to provide similar functionalities with its own parallel file systems.

The following table summarizes the characteristics of the file systems available in the DEEP-EST and DEEP-ER (SDV) systems:

Mount Point	User can write/read to/from	Cluster	Type	Global / Local	SW Version	Stripe Pattern Details	Maximum Measured Performance (see footnotes)	Description	Other
/p/home	/p/home/jusers/$USER	SDV, DEEP-EST	GPFS exported via NFS	Global				Home directory; used only for configuration files.
/p/project	/p/project/cdeep	SDV, DEEP-EST	GPFS exported via NFS	Global				Project directory; GPFS main storage file system; not suitable for performance relevant applications or benchmarking
/work	/work/cdeep	DEEP-EST*	BeeGFS	Global	BeeGFS 7.1.2			Work file system	*Also available in the SDV but only through 1 Gig network connection
/scratch	/scratch	DEEP-EST	xfs local partition	Local*				Scratch file system for temporary data. Will be cleaned up after job finishes!	*Recommended to use instead of /tmp for storing temporary files
/nvme/scratch	/nvme/scratch	DAM partition	local SSD (xfs)	Local*				Scratch file system for temporary data. Will be cleaned up after job finishes!	*1.5 TB Intel Optane SSD Data Center (DC) P4800X (NVMe PCIe3 x4, 2.5”, 3D XPoint))
/nvme/scratch2	/nvme/scratch2	DAM partition	local SSD (ext4)	Local*				Scratch file system for temporary data. Will be cleaned up after job finishes!	*1.5 TB Intel Optane SSD Data Center (DC) P4800X (NVMe PCIe3 x4, 2.5”, 3D XPoint))
/pmem/scratch	/pmem/scratch	DAM partition (odd nodes)	DCPMM in appdirect mode	Local*			2.2 GB/s simple dd test in dp-dam01		*3 TB in dp-dam01, 2 TB in dp-dam[03,05,07,09,11,13,15] Intel Optane DC Persistent Memory (DCPMM) 256GB DIMMs based on Intel’s 3D XPoint non-volatile memory technology
/sdv-work	/sdv-work/cdeep/$USER	SDV (deeper-sdv nodes via EXTOLL, KNL and ml-gpu via GbE only), DEEP-EST (1 GbE only)	BeeGFS	Global	BeeGFS 7.1.2	Type: RAID0, Chunksize: 512K, Number of storage targets: desired: 4	1831.85 MiB/s write, 1308.62 MiB/s read 15202 ops/s create, 5111 ops/s remove*	Work file system	*Test results and parameters used stored in JUBE: user@deep $ cd /usr/local/deep-er/sdv-benchmarks/synthetic/ior user@deep $ jube2 result benchmarks user@deep $ cd /usr/local/deep-er/sdv-benchmarks/synthetic/mdtest user@deep $ jube2 result benchmarks
/nvme	/nvme/tmp	SDV	NVMe device	Local	BeeGFS 7.1.2	Block size: 4K	1145 MiB/s write, 3108 MiB/s read 139148 ops/s create, 62587 ops/s remove*	1 NVMe device available at each SDV compute node	*Test results and parameters used stored in JUBE: user@deep $ cd /usr/local/deep-er/sdv-benchmarks/synthetic/ior user@deep $ jube2 result benchmarks user@deep $ cd /usr/local/deep-er/sdv-benchmarks/synthetic/mdtest user@deep $ jube2 result benchmarks
/mnt/beeond	/mnt/beeond	SDV	BeeGFS On Demand running on the NVMe	Local	BeeGFS 7.1.2	Block size: 512K	1130 MiB/s write, 2447 MiB/s read 12511 ops/s create, 18424 ops/s remove*	1 BeeOND instance running on each NVMe device	*Test results and parameters used stored in JUBE: user@deep $ cd /usr/local/deep-er/sdv-benchmarks/synthetic/ior user@deep $ jube2 result benchmarks user@deep $ cd /usr/local/deep-er/sdv-benchmarks/synthetic/mdtest user@deep $ jube2 result benchmarks

Stripe Pattern Details

It is possible to query this information from the deep login node, for instance:

manzano@deep $ fhgfs-ctl --getentryinfo /work/manzano
Path: /manzano
Mount: /work
EntryID: 1D-53BA4FF8-3BD3
Metadata node: deep-fs02 [ID: 15315]
Stripe pattern details:
+ Type: RAID0
+ Chunksize: 512K
+ Number of storage targets: desired: 4

manzano@deep $ beegfs-ctl --getentryinfo /sdv-work/manzano
Path: /manzano
Mount: /sdv-work
EntryID: 0-565C499C-1
Metadata node: deeper-fs01 [ID: 1]
Stripe pattern details:
+ Type: RAID0
+ Chunksize: 512K
+ Number of storage targets: desired: 4

Or like this:

manzano@deep $ stat -f /work/manzano
  File: "/work/manzano"
    ID: 0        Namelen: 255     Type: fhgfs
Block size: 524288     Fundamental block size: 524288
Blocks: Total: 120178676  Free: 65045470   Available: 65045470
Inodes: Total: 0          Free: 0

manzano@deep $ stat -f /sdv-work/manzano
  File: "/sdv-work/manzano"
    ID: 0        Namelen: 255     Type: fhgfs
Block size: 524288     Fundamental block size: 524288
Blocks: Total: 120154793  Free: 110378947  Available: 110378947
Inodes: Total: 0          Free: 0

See ​http://www.beegfs.com/wiki/Striping for more information.

Additional infos

Detailed information on the BeeGFS Configuration can be found ​here.

Detailed information on the BeeOND Configuration can be found ​here.

Detailed information on the Storage Configuration can be found ​here.

Detailed information on the Storage Performance can be found ​here.

Notes

• dd test @dp-dam01 of the DCPMM in appdirect mode:

[root@dp-dam01 scratch]# dd if=/dev/zero of=./delme bs=4M count=1024 conv=sync
1024+0 records in
1024+0 records out
4294967296 bytes (4.3 GB) copied, 1.94668 s, 2.2 GB/s

• The /work file system which is available in the DEEP-EST prototype, is as well reachable from the nodes in the SDV (including KNLs and ml-gpu nodes) but through a slower connection of 1 Gig. The file system is therefore not suitable for benchmarking or I/O task intensive jobs from those nodes

• Performance tests (IOR and mdtest) reports are available in the BSCW under DEEP-ER → Work Packages (WPs) → WP4 → T4.5 - Performance measurement and evaluation of I/O software → Jülich DEEP Cluster → Benchmarking reports: ​https://bscw.zam.kfa-juelich.de/bscw/bscw.cgi/1382059

• Test results and parameters used are stored in JUBE:

user@deep $ cd /usr/local/deep-er/sdv-benchmarks/synthetic/ior
user@deep $ jube2 result benchmarks

user@deep $ cd /usr/local/deep-er/sdv-benchmarks/synthetic/mdtest
user@deep $ jube2 result benchmarks