Dell PowerFlex 4.6.x Technical Overview

PDF

Storage virtual memory requirements

The following is a list of PowerFlex storage virtual memory requirements.

Storage virtual memory requirements for Dell PowerEdge servers (R660/R760), storage only nodes include non-volatile random-access memory (SDPM), and storage only nodes include non-volatile random-access memory (NVDIMM) Dell PowerEdge servers (R650/R750).

Table 1. Storage virtual memory (SVM) requirementsThe following table describes the storage virtual memory requirements.
Component	Memory allocation rules
Memory	PowerFlex RAM (GiB) SDS FG: 10 GiB + 550 MiB per 1 TiB device size + 100 MiB per device SDS MG: 5 GiB + 335 MiB per 1TiB device size MD Read Cache: Total_Drive_Capacity_in_TiB * 8% MDM: 6.5 GiB CloudLink: 4 GiB Operating System: 1 GiB LIA: 0.35 GiB SDC: 6 GiB SDR: 8586 MiB + 550 MiB * (max remote PDs count) NOTE:SDR SVM support is only available as part of a PowerFlex appliance or the PowerFlex rack managed by PowerFlex Manager. SDR memory requirement for hyperconverged systems (SVMs): 12GB NOTE:For certain system configurations, the SDS needs to be configured to be able to use more than one non-uniform memory access (NUMA) domain. The SDS is configured by default to have affinity to socket 0 in a server or VM, and by default, the SDS is connected only to NUMA 0, and only has access to the memory in NUMA 0. If NUMA 0 (usually half the total memory) is less than the memory required by the SDS, you need to allow the SDS access to the memory in the other NUMA. Non-volatile RAM SDPM: Node storage capacity less than 41.9 TB: SDPM size 32 GiB Node storage capacity 41.9 to 91 TiB: SDPM size 64 GiB Node storage capacity greater than 91 TiB and less than 112 TiB: SDPM size 96 GiB or non-volatile RAM (NVDIMM): Node storage capacity less than 49 TiB: NVDIMM size 32 GiB Node storage capacity 49 to 96 TiB: NVDIMM size 64 GiB Node storage capacity greater than 96 TiB: NVDIMM size 96 GiB NOTE:PowerFlex does not support using swap on Storage only and SVM nodes. This should be the default setting. Using swap with PowerFlex has the following impact: Slows the performance of the system. Impacts I/O on the system disks which are used for swap. This may cause other functions that use the system disk to have slow I/Os such as: MDM repository-related operations PowerFlex logs written to disk Disable swap and set vm.overcommit to the following on nodes that have MDM and SDS processes on them in the in/etc/sysctl.conf file: vm.overcommit_memory=2 vm.overcommit_ratio=100
Base SVM	350 MiB
Tiebreaker MDM	50 MiB