PowerScale OneFS 9.4.0.0 Technical Specifications Guide

PDF

OneFS software module guidelines

This section presents guidelines for configuring OneFS software modules.

For assistance, contact your PowerScale account representative or PowerScale Technical Support.

Table 1. OneFS software module specifications
Item	OneFS 9.4.0.0	Description
Anti-virus: file size	ICAP: 2 GB CAVA: 16 TiB	The recommended and maximum allowed file size limit for anti-virus scans. Evaluate the workflow and workloads for your cluster to determine the value that works best for your environment.
Anti-virus: scan report entries	10000	The maximum number of anti-virus scan and threat reports that can be fetched from a configuration at a given time. Reports beyond the limit can be fetched by configuring the offset parameter for the configuration.
Audit: CEE servers	1 (max) audit server per node 252 (max) audit servers per cluster	OneFS must ping all the Common Event Enabler (CEE) servers within a single heartbeat window. The number of servers that can be contacted and that can respond during the window is estimated to be 252. The network topology and cluster bandwidth might require a lower limit.
Audit: Events forwarded to CEE	4500 events per second	The sustained number of audit events, per second, that can be forwarded to a CEE server. This limit might be higher in some circumstances, depending on the workload, the type of node, and the CEE server configuration.
Audit: log expiration	User configurable	Audit logs can be autodeleted from the system by specifying a retention period. Minimum retention period that can be specified is 1 day. Logs can also be deleted manually by specifying a delete-before date. NOTE: Logs are not deleted until all the contained events have been forwarded to a CEE server.
Audit: log file size	1 GB	The size limit for audit log files. When a log file reaches the maximum size, the log file is closed and a new log file is created. Old log files can be deleted from the cluster using manual or auto-delete methods.
Audit: maximum size of an audit event	65535 bytes	This is the maximum supported size for an audit event. If the size of an audit event is greater than 65,535, that log event is discarded and the file access operation that caused the event fails.
CloudPools: account name	768 characters	The maximum length for a CloudPool account name.
CloudPools: account username	Service provider sets this limit	The maximum length for a CloudPool account . This limit is set by the service provider. Check with your cloud provider for more information.
CloudPools: account password	255 characters	The maximum length for a CloudPool account password.
CloudPools: pool name	768 characters	The maximum length for a CloudPool name
CloudPools: vendor name	2048 characters	The maximum length for a CloudPool vendor name.
CloudPools: description	4096 characters	The maximum length for a CloudPool description.
CloudPools: accounts to tier to	256 accounts 30 active accounts	The maximum number of accounts that a CloudPool account can tier to. The number of accounts that can be active is limited by the maximum number of file pool policies.
CloudPools: containers in cloud	The service provider sets this limit	The maximum number of containers in the cloud. This limit is set by the service provider. Check with your cloud provider for more information.
CloudPools: cloud container size	The service provider sets this limit	The size of the cloud container. The service provider sets this limit. Check with your cloud provider for more information.
CloudPools: storage size per CloudPool account	The service provider sets this limit	The storage size for a CloudPool account. The service provider sets this limit. Check with your cloud provider for more information.
CloudPools: file size tiered to cloud	4.398 TB (4 TiB)	The size of files that can be archived to the cloud and retrieved from the cloud. The service provider sets this limit. Check with your cloud provider for more information.
CloudPools: proxy limits	Proxy name: 1024 characters Proxy hostname: 1024 characters Proxy username: 1024 characters Proxy password: 256 characters	The maximum lengths for a CloudPool proxy name, hostname, username, and password.
Datamover Jobs	1000	The limit for concurrently executed datamover jobs. Datamover can only replicate one same dataset to at most ~126 target locations on the same cluster, else the job can not complete successfully.
Datamover Policies	1000	The recommended limit for datamover policies. Evaluate the workflow and workloads for your cluster to determine the value that works best for your environment. The maximum limit has not been established.
Datamover Accounts	1000	The recommended limit datamover accounts. Evaluate the workflow and workloads for your cluster to determine the value that works best for your environment. The maximum limit has not been established.
Datamover Datasets	1000	The recommended limit for datamover datasets. Evaluate the workflow and workloads for your cluster to determine the value that works best for your environment. The maximum limit has not been established.
File pool policies: AND and OR conditions	3 ORs and 5 ANDs	A file pool policy can have 3 OR disjunctions, and each term joined by the ORs can contain at most 5 ANDs. For example: (A and B and C and D and E) or (F and G and H and I and J) or (K and L and M and N and O).
File pool policies: number of file pool policies per cluster	256	The recommended limit for file policies per cluster. Exceeding this limit might negatively affect the cluster performance and client connections. Evaluate the workflow and workloads for your cluster to determine the value that works best for your environment.
InsightIQ 4.1.1 and later: monitored clusters	8 clusters	The number of clusters that InsightIQ 4.1.1 and later can monitor. If you are running a different version of InsightIQ, see the release notes for the version of InsightIQ that you are running.
InsightIQ 4.1.1 and later: monitored nodes	80 nodes for one cluster; 150 nodes for multiple clusters	The number of nodes for one cluster or multiple clusters that InsightIQ 4.1.1 and later can monitor. If you are running a different version of InsightIQ, see the release notes for the version of InsightIQ that you are running.
Job Engine: concurrent jobs	3	The number of concurrent jobs that the job engine can run. However, the job Exclusion Sets (restripe or marking) determine which jobs can be run simultaneously. Concurrent job execution is also governed by job priority and overall cluster health. For more information, see the OneFS Job Engine White Paper.
MFTv3: concurrent downloads	1	Maximum number of concurrent downloads through EMC Secure Remote Services (ESRS).
SmartDedupe: block size	8 KB	SmartDedupe works on file system blocks that are 8 KB.
SmartDedupe: maximum paths per job	10	The recommended limit for paths per job for SmartDedupe. Exceeding this limit might negatively affect the cluster performance and client connections. Evaluate the workflow and workloads for your cluster to determine the value that works best for your environment.
SmartDedupe: minimum file size	32 KB	The minimum file size that SmartDedupe can process. SmartDedupe will not deduplicate files that are smaller than 32 KB.
SmartDedupe: shadow stores	32,000	Each shadow store can have 32,000 pointers. This limit is imposed by the kernel. The OneFS shadow store is a metadata structure that references physical blocks to decrease the physical storage that is required to store data, which maximizes storage efficiency.
SmartPools: Tiers	5	The recommended limit for SmartPools tiers. Exceeding this limit might negatively affect the cluster performance and client connections. Evaluate the workflow and workloads for your cluster to determine the value that works best for your environment.
SmartQuotas: directory depth	509	The maximum limit for directory depths for SmartQuotas. Directory depths deeper than 275 directories might negatively affect system performance.
SmartQuotas: number of quotas per cluster	500,000	The recommended limit for quotas per cluster. The maximum number of quotas per cluster has not been established. Exceeding this recommended limit might negatively affect the cluster performance and client connections. Listing of quotas in the WebUI is expected to take time. For assistance, contact your PowerScale account representative or Dell Technologies Technical Support.
SnapshotIQ: directory depth	509	The maximum limit for directory depths for SnapshotIQ. Directory depths deeper than 275 directories might negatively affect system performance.
SnapshotIQ: number of snapshots	20,000	The limit for snapshots per cluster
SnapshotIQ: Number of writable snapshots	Default: 30 Maximum supported: 2048 with limitations*	The limit for writable snapshots per cluster. Limitations: Do not delete all writable snapshots at the same time which can lead to filling up of Job Engine queue.
SyncIQ: defined policies	1,000	The recommended limit for defined SyncIQ policies. The maximum limit of defined policies has not been established. If the number of policies exceeds the recommended limit, you should keep in mind the following effects: SyncIQ is bound by the limit on the number of concurrently running policies. If many policies are running on schedules, the queue to run the jobs might become so large that OneFS can never complete all the jobs in the queue. Each policy represents a set of snapshots on the source and the destination clusters. More snapshots mean that more jobs must run to delete the snapshots, and the increase in the number of jobs can negatively affect the cluster performance.
SyncIQ: running policies	50 - for clusters with 4 or more nodes OR 4 * number of CPU cores per cluster - for clusters with 3 or fewer nodes	The recommended limit of running SyncIQ policies. For clusters with 3 or fewer nodes, the limit depends on the number of CPU cores per node. There can be one worker per CPU core, with each worker running 4 policies. The recommended limit for smaller clusters is: 4 * number of CPU cores per cluster. Exceeding this limit might negatively affect the cluster performance and client connections. Evaluate the workflow and workloads for your cluster to determine the value that works best for your environment.
SyncIQ: workers per node (policy setting)	3	The recommended limit for workers per node. Exceeding this limit might negatively affect the cluster performance and client connections. Evaluate the workflow and workloads for your cluster to determine the value that works best for your environment.
SyncIQ: workers per policy	40	The recommended limit for workers per policy. Exceeding this limit might negatively affect the cluster performance and client connections. Evaluate the workflow and workloads for your cluster to determine the value that works best for your environment.