gpfs:gpfs_sizing
Differences
This shows you the differences between two versions of the page.
| Both sides previous revision Previous revision Next revision | Previous revision | ||
|
gpfs:gpfs_sizing [2021/04/30 22:42] manu |
gpfs:gpfs_sizing [2025/11/17 11:04] (current) manu [Features in IBM Storage Scale editions] |
||
|---|---|---|---|
| Line 1: | Line 1: | ||
| ====== Spectrum Scale sizing ====== | ====== Spectrum Scale sizing ====== | ||
| + | ===== VM sizing for IBM Storage Scale ===== | ||
| + | |||
| + | |||
| + | Sizing Summary for a Small Cluster | ||
| + | |||
| + | A common starting point for a small, moderately performing virtual Storage Scale cluster might be: | ||
| + | * Minimum Nodes: 3 (for quorum and basic replication) | ||
| + | * Per NSD Server VM: | ||
| + | <code> | ||
| + | vCPU: 8 vCPUs | ||
| + | vRAM: 16 GB | ||
| + | Network: 2 x 25 Gbps vNICs (bonded for I/O) | ||
| + | </code> | ||
| + | * Per Protocol node: | ||
| + | <code> | ||
| + | vCPU: 1 vCPUs | ||
| + | vRAM: 64 GB or 128 GB if NFS ans SMB is used | ||
| + | Network: 10 Gbps | ||
| + | </code> | ||
| + | * Storage: Virtual disks backed by high-speed SAN LUNs (or local disks on the hypervisor host if using FPO). | ||
| + | * Best Practice: Use physical mode Raw Device Mapping (RDM) or the virtualization platform's equivalent for best I/O performance and direct control over the LUNs from the Storage Scale VMs. | ||
| + | |||
| + | https://community.ibm.com/community/user/blogs/ramesh-krishnamaneni/2025/09/26/optimizing-ibm-storage-scale-in-ibm-cloud-vsi-vs-b | ||
| + | |||
| + | ===== Features in IBM Storage Scale editions ===== | ||
| + | |||
| + | ^ Feature ^ Data Access ^ Data Management1 ^ Erasure Code Edition ^ | ||
| + | | Multi-protocol scalable file service with simultaneous access to a common set of data | ✓ | ✓ | ✓ | | ||
| + | | Facilitate data access with a global namespace, massively scalable file system, quotas and snapshots, data integrity and availability, and filesets | ✓ | ✓ | ✓ | | ||
| + | | Simplify management with GUI | ✓ | ✓ | ✓ | | ||
| + | | Improved efficiency with QoS and compression | ✓ | ✓ | ✓ | | ||
| + | | Create optimized tiered storage pools based on performance, locality, or cost | ✓ | ✓ | ✓ | | ||
| + | | Simplify data management with Information Lifecycle Management (ILM) tools that include policy based data placement and migration | ✓ | ✓ | ✓ | | ||
| + | | Enable worldwide data access using AFM asynchronous replication | ✓ | ✓ | ✓ | | ||
| + | | Asynchronous multi-site Disaster Recovery | | ✓ | ✓ | | ||
| + | | Multi-site replication with AFM to cloud object storage | | ✓ | ✓ | | ||
| + | | Protect data with native software encryption and secure erase, NIST compliant and FIPS certified | | ✓ | ✓ | | ||
| + | | File audit logging | | ✓ | ✓ | | ||
| + | | Watch folder | | ✓ | ✓ | | ||
| + | | Erasure coding | ESS only | ESS only | ✓ | | ||
| + | |||
| + | https://www.ibm.com/docs/en/storage-scale/5.2.3?topic=overview-storage-scale-product-editions | ||
| ===== Filesystem Block size, best practice ===== | ===== Filesystem Block size, best practice ===== | ||
| + | |||
| + | Typically, metadata is between 1 and 5% of the filesystem space, but this can vary. | ||
| Depending on usage, you can have different **block size** | Depending on usage, you can have different **block size** | ||
| Line 53: | Line 97: | ||
| During NSD creation alternate node position for access to NSD. | During NSD creation alternate node position for access to NSD. | ||
| If the first node is used as first node in NSD definition, then it 'll be only use and you'll reach performance problems | If the first node is used as first node in NSD definition, then it 'll be only use and you'll reach performance problems | ||
| - | I have a question regarding your NSD configuration below. Starting on NSD 'data_nsd043' up to NSD 'data_nsd106' the primary NSD server role rotated. | + | |
| - | Then you stopped to rotate primary server at NSD 'data_nsd107' and configured 'gpfs1' as your primary server. Which means every NSD task ( read/ write ... ) has to be handled by NSD server 'gpfs1' as long as he is reachable. | + | If only the first node is selected as first into NSD definition means every NSD task ( read/ write ... ) has to be handled by NSD server 'gpfs1' as long as he is reachable. |
| - | Such a configuration could cause a overload situation on the affected server. So my recommendation is to change it and to rotate it in the same manner like did up to NSD 'data_nsd106'! | + | Such a configuration could cause a overload situation on the affected server. |
| - | The NSD server sequence can be adjusted via command 'mmchnsd' ( see below ): | + | |
| + | The NSD server sequence can be adjusted online via command **mmchnsd** ( see below ): | ||
| https://www.ibm.com/docs/en/spectrum-scale/5.0.5?topic=disks-changing-your-nsd-configuration | https://www.ibm.com/docs/en/spectrum-scale/5.0.5?topic=disks-changing-your-nsd-configuration | ||
| - | <cli> | + | Ex: |
| + | <cli prompt='#'> | ||
| + | # mmchnsd "data_nsd043:gpfs03.gpfsint.labo,gpfs04.gpfsint.labo,gpfs01.gpfsint.labo,gpfs02.gpfsint.labo" | ||
| + | </cli> | ||
| + | |||
| + | Maybe easier with a description file | ||
| + | <cli prompt='#'> | ||
| + | # mmlsnsd -X | ||
| File system Disk name NSD volume ID NSD servers | File system Disk name NSD volume ID NSD servers | ||
| ------------------------------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------ | ||
gpfs/gpfs_sizing.1619815351.txt.gz · Last modified: 2021/04/30 22:42 by manu
Page Tools
Except where otherwise noted, content on this wiki is licensed under the following license: GNU Free Documentation License 1.3
