«EMC® VNX™ Series Release 7.0 Managing Volumes and File Systems on VNX™ Manually P/N 300-011-808 REV A01 EMC Corporation Corporate Headquarters: ...»
◆ MirrorView/Synchronous on page 32 Quotas To ensure that file systems do not become full, you can impose quota limits on users and groups that create files on directory trees. You can set a hard quota limit on user, group, or directory tree quotas to prevent allocation of all the space in the file system. When the hard quota limit is reached, the system denies user requests to save additional files and notifies the administrator that the hard quota limit has been reached. In this case, existing files can be read but action must be taken either by the user or administrator to delete files from the file system or increase the hard quota limit to allow saving of additional files.
To avoid degradation of file system performance, set the hard quota limit between 80 and 85 percent of the total file system space. In addition to setting the hard quota limit, set a lower soft quota limit so that the administrator is notified when the hard quota limit is being approached.
For example, to prevent a file system that contains 100 GB of storage from filling up, you can set a soft quota limit of 80 GB and a hard quota limit of 85 GB by using user, group, or directory tree quotas. When used space in the file system reaches 80 GB, the administrator is notified that the soft limit is reached. When used space totals 85 GB, the system denies user requests to save additional files, and the administrator is notified that the hard quota limit is reached.
Using Quotas on VNX provides detailed information on quotas and how to set up user, group, or directory tree quotas.
TimeFinder/FS If you plan to create multiple copies of your PFS, plan for that number of BCVs. For example, from one PFS, you can create 10 copies. Therefore, plan for 10 BCVs, not one.
TimeFinder/FS uses the physical disk, not the logical volume, when it creates BCV copies.
The copy is done track-by-track, so unused capacity is carried over to the BCVs.
Volumes used for BCVs should be of the same size as the standard volume.
Using TimeFinder/FS, NearCopy, and FarCopy on VNX provides additional information on TimeFinder/FS.
File-level retention File systems can only be enabled with file-level retention (FLR) capability at creation time.
When the file system is created and enabled for FLR, it is persistently marked as an FLR file system and the FLR setting cannot be changed. After a file system is created and FLR is enabled, an administrator can apply FLR protection to individual files. Files in the FLR (locked) state can be stored with retention dates, which prohibit the deletion of the file until expiration.
Using VNX File-Level Retention provides more information about FLR storage and FLR file system behavior.
Planning considerations 31Concepts
SRDF All file systems on the Data Mover must be built on SRDF volumes. Using SRDF/S with VNX for Disaster Recovery describes SRDF/S and Using SRDF/A with VNX describes SRDF/A.
If you use the AVM feature to create the file systems, specify the symm_std_rdf_src storage pool. This storage pool directs AVM to allocate space from volumes configured during installation for remote mirroring by using SRDF.
Automatic file system extension cannot be used for any file system that is part of a Remote Data Facility (RDF) configuration.
Note: Do not use the nas_fs command with the -auto_extend option for file systems associated with RDF configurations. Doing so generates the error message: Error 4121: operation not supported for file systems of type SRDF.
MPFS File systems mounted read-only are not acknowledged by clients that use MPFS, and thereby allow clients to write to the file system.
You cannot enable MPFS access to file systems with a stripe depth of less than 32 KB. Using VNX Multi-Path File System provides additional information on MPFS.
VNX Replicator/SnapSure By using VNX Replicator, you can enable automatic file system extension on the source file system. When the source file system hits its high water mark (HWM), the destination file system automatically extends first. Then the source file system automatically extends. If the extension of the destination file system succeeds but the source file system extension fails, the file systems differ in size, which causes replication failure. Use the nas_fs -xtend fs_name -option src_only command to manually adjust the size of the source file system.
Using VNX Replicator contains instructions to recover from this situation. Managing Volumes and File Systems with VNX AVM provides information on automatic file system extension.
There must be sufficient file system space and disk storage available to support VNX Replicator and SnapSure operations. To review the entire file system size, use the nas_fs
-list command. To calculate the SavVol file size, use the nas_disk -size command. The VNX Command Line Interface Reference for File provides a detailed synopsis of the commands associated with SnapSure and VNX Replicator.
MirrorView/Synchronous All file systems on the Data Mover must be built on MirrorView/Synchronous LUNs. Using MirrorView/Synchronous with VNX for Disaster Recovery provides detailed information on MirrorView/Synchronous.
If you use the AVM feature to create the file systems, you must use the appropriate MirrorView AVM storage pools for your RAID configuration. Managing Volumes and File Systems with VNX AVM provides a list of the MirrorView AVM storage pools.
32 Managing Volumes and File Systems on VNX Manually 7.0
The tasks to configure volumes manually are:
◆ Determine storage availability on page 34 ◆ Create volumes on page 35 ◆ Provide storage from a VNX or legacy CLARiiON system to a gateway system on page 38 ◆ Add disk volumes to an integrated system on page 41
Determine storage availability When reserving disk volumes for manual volume management or for use with AVM user-defined storage pools, reserve disk volumes in SP-balanced pairs of the same RAID type, disk count, and size. Although you can use all available disk volumes with AVM user-defined storage pools, its performance might be impacted if the volumes are not configured this way.
Before you create a new volume, identify the unused disk space. If a disk is unused, its space is available for volume and file system creation. Determine storage availability on the system by identifying the unused disks.
You can create three types of volumes:
◆ Slice volume ◆ Stripe volume ◆ Metavolume Each volume type provides different benefits to satisfy storage requirements. Volumes on page 20 provides detailed information on volume types. Volume configuration guidelines on page 29 lists common volume configurations and considerations associated with each volume type.
List volumes Action
To list all volumes on a system, type:
$ nas_volume -list Output
This is a partial listing of the volume table that is displayed as the volume list:
Note You can also use the nas_slice -list command to list only slice volumes.
id — ID of the volume (assigned automatically) inuse — Indicates whether the volume is in use by a file system; y indicates yes, n indicates no type — Type of volume acl — Access control value assigned to the volume name — Name of the volume
When creating a stripe volume, if you do not specify a name for the stripe volume, a default name is assigned. Stripe volume configuration considerations on page 30 provides more information.
To create a stripe volume, use this command syntax:
$ nas_volume -name name -create -Stripe stripe_size [volume_name,...]
name = name of the stripe volume stripe_size = size of the stripe volume in megabytes volume_name = names of the volumes separated by commas
To create a stripe volume called stv1, type:
$ nas_volume -name stv1 -create -Stripe 32768 d10,d12,d13,d15 Output
Create a metavolume When creating a metavolume, if you do not specify a name for the metavolume, a default name is assigned.
To combine volumes into a metavolume, use the volume_name option consecutively in the command syntax.
To create a metavolume from a stripe volume, use this command syntax:
$ nas_volume -name name -create -Meta [volume_name,...]
name = name of the stripe volume volume_name = names of the metavolumes separated by commas
To create metavolumes named slv1, slv2, and slv3 on disk volume d7, type:
$ nas_volume -name mtv1 -create -Meta slv1,slv2,slv3
Provide storage from a VNX or legacy CLARiiON system to a gateway system Before you begin If a VNX for file gateway system is attached to a VNX for block storage system and you want to add disk volumes to the configuration, use this procedure.
EMC VNX for file network servers that are gateway network-attached storage systems and
that connect to EMC Symmetrix and VNX for block storage systems are:
◆ VNX VG2 ◆ VNX VG8 A VNX for file gateway system stores data on VNX for block user LUNs or Symmetrix hypervolumes. If the user LUNs or hypervolumes are not configured correctly on the storage system, AVM and the Unisphere for File software cannot be used to manage the storage.
Typically, an EMC Customer Support Representative does the initial setup of disk volumes on these gateway storage systems.
Use the procedures that follow to add disk volumes to a configuration:
1. Use the Unisphere software or the VNX for block CLI to create user LUNs.
2. Use either the Unisphere software or the VNX for file CLI to make the new user LUNs available to the VNX for file as disk volumes.
Create user LUNs before you create file systems.
To add user LUNs, you must be familiar with the following:
◆ Unisphere software or the VNX for block CLI.
◆ Process of creating RAID groups and user LUNs for the VNX for file volumes.
The documentation for Unisphere and VNX for block CLI, describes how to create RAID groups and user LUNs.
1. Create RAID groups and user LUNs as needed for VNX for file volumes. Ensure that you add the LUNs to the VNX for file gateway system's storage group and that you set
the HLU to 16 or higher:
• Always create the user LUNs in balanced pairs, one owned by SP A and one owned by SP B. The paired LUNs must be the same size.
• FC or SAS disks must be configured as RAID 1/0, RAID 5, or RAID 6. The paired LUNs do not have to be in the same RAID group but should be of the same RAID type. RAID groups and storage characteristics on page 40 lists the valid RAID group and storage system combinations. Gateway models use the same combinations as the NS-80 (for CX3™ series storage systems) or the NS-960 (for CX4™ series storage systems).
• SATA disks must be configured as RAID 1/0, RAID 5, or RAID 6. All LUNs in a RAID group must belong to the same SP; create pairs by using LUNs from two RAID groups.
RAID groups and storage characteristics on page 40 lists the valid RAID group and storage system combinations. Gateway models use the same combinations as the NS-80 (for CX3 series storage systems) or the NS-960 (for CX4 series storage systems).
• The host LUN identifier (HLU) is greater than or equal to 16 for user LUNs.
Use these settings when creating RAID group user LUNs:
• RAID Type: RAID 1, RAID 5, or RAID 6 for FC disks and RAID 3, RAID 5, or RAID 6 for ATA disk • LUN ID: Select the first available value • Rebuild Priority: ASAP • Verify Priority: ASAP • Enable Read Cache: Selected • Enable Write Cache: Selected • Enable Auto Assign: Cleared (off) • Number of LUNs to Bind: 2 • Alignment Offset: 0 • LUN size: Must not exceed 14 TB Note: If you create 4+1 RAID 3 LUNs, the Number of LUNs to Bind value is 1.
2. Perform one of these steps to make the new user LUNs available to the VNX for file:
Using the Unisphere software:
a. Select Storage ➤ Storage Configuration ➤ File Systems.
b. From the task list, select File Storage ➤ Rescan Storage Systems.
Add disk volumes to an integrated system Configure unused or new disk devices on a VNX for block storage system by using the Disk Provisioning Wizard for File. This wizard is available only for integrated VNX for file models (NX4 and NS non-gateway systems excluding NS80), including Fibre Channel-enabled models, attached to a single VNX for block storage system.
Note: For VNX systems, Advanced Data Service Policy features such as FAST and compression are supported on pool-based LUNs only. They are not supported on RAID-based LUNs.
To open the Disk Provisioning Wizard for File in the Unisphere software:
1. Select Storage ➤ Storage Configuration ➤ Storage Pools.
2. From the task list, select Wizards ➤ Disk Provisioning Wizard for File.
Note: To use the Disk Provisioning Wizard for File, you must log in to Unisphere by using the global sysadmin user account or by using a user account which has privileges to manage storage.
An alternative to the Disk Provisioning Wizard for File is available by using the VNX for file CLI at /nas/sbin/setup_clariion. This alternative is not available for unified VNX systems.
The script performs the following actions:
◆ Provisions the disks on integrated (non-Performance) VNX for block storage systems when there are unbound disks to configure. This script binds the data LUNs on the xPEs and DAEs, and makes them accessible to the Data Movers.
◆ Ensures that your RAID groups and LUN settings are appropriate for your VNX for file server configuration.
The Unisphere for File software supports only the array templates for EMC CLARiiON CX™ and CX3 storage systems. CX4 and VNX systems must use the User_Defined mode with the /nas/sbin/setup_clariion CLI script.
Add disk volumes to an integrated system 41Configuring Volumes
The setup_clariion script allows you to configure VNX for block storage systems on a shelf-by-shelf basis by using predefined configuration templates. For each enclosure (xPE or DAE), the script examines your specific hardware configuration and gives you a choice of appropriate templates. You can mix combinations of RAID configurations on the same storage system. The script then combines the shelf templates into a custom, User_Defined array template for each VNX for block system, and then configures your array.
42 Managing Volumes and File Systems on VNX Manually 7.0
The tasks to configure file systems are:
◆ Create a file system on page 44 ◆ (Optional) Create a mount point on page 45 ◆ Mount a file system on page 46 ◆ (Optional) Create an NMFS on page 47
Create a file system File systems can be created only on nonroot metavolumes not in use. The metavolume must be at least 2 megabytes to accommodate a file system.
A file system name must be unique on a particular system. Name length is limited to 240 bytes (represented as 240 ASCII characters or a variable number of Unicode multibyte characters). A file system name can include upper and lowercase letters, numbers, hyphens (-), underscores (_), and periods (.). Alphanumeric names are accepted.
A file system name cannot:
◆ Begin with a hyphen, period, or the word root ◆ Include a blank character or colon (:) ◆ Be a single integer ◆ Be comprised entirely of integers