NetScaler Management and Analytics System (MAS) storage requirement is determined based on your NetScaler MAS sizing estimation. If your NetScaler MAS storage requirement exceeds the default disk space (120 gigabytes), you can attach an additional disk. You can attach additional disk in both single-server and high availability deployments. For NetScaler MAS servers in a high availability setup, you need to add an additional disk for each node.
- You can attach additional disks of capacity greater than 2 terabytes.
- Citrix recommends using solid-state drive (SSD) technology for NetScaler MAS deployments.
- You must estimate storage requirements and attach an additional disk to the server at the time of initial deployment of NetScaler MAS.
- For a NetScaler MAS high availability deployment, you must attach an additional disk to each node.
NetScaler MAS now provides NetScaler MAS Disk Partition Tool, a new command line tool. The functionalities of this tool are described in detail as follows:
1. Using the tool, you can create partitions in the newly added additional disk. The size of the additional disk can be lesser or greater than 2 terabytes.
2. You can also resize existing disks to a maximum capacity of 2 terabytes using this tool.
Note: It is still not possible to resize existing disks beyond 2 terabytes without losing data due to a known limitation on the platform. To create a storage capacity greater than 2 terabytes, you must remove the existing partitions and create new partitions using this new tool.
3. Using this new tool, you can perform any partition action on the disk explicitly. The tool provides you with clear visibility and control over the disk and the associated data.
Note: You can only use this tool on the additional disk that you have attached to the NetScaler MAS server. You cannot create partitions in the primary disk using this tool.
This document explains the following scenarios about attaching a new additional disk, creating partitions, and resizing the additional disks:
- Attaching a new additional disk
- Launching the disk partitioning tool
- Creating partitions in the new additional disk
- Resizing the existing additional disk
- Removing partitions on the additional disk
To attach an additional disk in a standalone NetScaler MAS
For a NetScaler MAS single-server deployment, you can attach only one disk to the server in addition to the default disk.
Perform the following steps to attach a disk to the virtual machine:
- Shut down the NetScaler MAS virtual machine.
- In the hypervisor, attach an additional disk of the required disk size to NetScaler MAS virtual machine. For example, you can attach an additional disk of 2 terabytes capacity to increase disk space from 120 gigabytes. The newly attached disk stores the database data, and NetScaler MAS log files. The existing 120 gigabytes disk space stores the core files, operating system log files, and so on.
- Start the NetScaler MAS virtual machine.
To launch the disk partition tool
- Open an SSH connection to the NetScaler by using an SSH client, such as PuTTY.
- Log on to the NetScaler by using the administrator credentials.
- Switch to the shell prompt and type:
For NetScaler MAS in high availability deployment, you must launch the tool in both nodes and create or resize partitions after attaching disks to the respective virtual machines.
To create partitions in the new additional disk
The "create" command is used to create new partitions whenever a new secondary disk is added. You can also use this command to create new partitions on an existing secondary disk after the existing partitions are deleted using the "remove" command.
Note: There is no 2 terabytes size limitation while creating new partitions with the disk partition tool. The tool can create partitions larger than 2 terabytes. While partitioning the disk, a swap partition of size 32 gigabytes is automatically added. The primary partition then uses all the remaining space on the disk.
Once the command is executed, it creates a GUID partition table (GPT) partition scheme and creates the 32 gigabytes swap partition and data partition to use rest of the space. A new file system is then created on the primary partition.
Note: This process can take a few seconds, and you must not interrupt the process.
Once the create command completes, the virtual machine is automatically restarted for the new partition to get mounted.
After the restart, the new partition is mounted at /var/mps.
The swap partition added shows up as swap space in the output of the "create" command.
The tool restarts the virtual machine after you have created the partition. You can also manually restart the virtual machine by using the "reboot" command.
To resize the partitions in the existing additional disk
You can use the "resize" command to resize the attached (secondary) disk. You can resize a disk that has a master boot record (MBR) or GPT scheme and is less than 2 terabytes in size to a maximum of 2 terabytes.
- The "resize" command is designed to function without losing any existing data, but Citrix recommends that you back up critical data in this disk to external storage before attempting the resize. Data backup is helpful in cases where the disk data can get corrupted during the resize operation. For more information on how to take a backup of the data, see To create a backup file of the disk data.
- Ensure to increase the disk space in increments of 100 gigabytes of space while resizing the partitions. Such an incremental increase ensures that you would not have to resize more frequently.
The "resize" command checks for all preconditions and proceeds if all preconditions are met and that you have given consent to resizing. It stops all the processes accessing the disk which includes the NetScaler MAS subsystems, PostgreSQL DB processes, and NetScaler MAS monitor process. Once the processes are stopped, the disk is unmounted to prepare it for resizing. The resizing is done by extending the partition to occupy all available space and then growing the file system. If a swap partition exists on the disk, it is deleted and recreated at the end of the disk after resizing. The swap partition is discussed in the Create command section of the document.
Note: The "growing file system" process can take some to complete and take care that you do not interrupt the process while it is in progress. The tool restarts the virtual machine after you have resized the partition. You can also manually restart the virtual machine by using the "reboot" command.
All the intermediate steps in the resize process (stopping applications, resizing disk, growing filesystem) are shown on the console and once the process completes the following message is seen.
After rebooting, the increase in size can be observed using the "df" command.
To remove the partitions in the additional disk
An existing partition on the secondary disk can be resized only up to 2 terabytes due to a known limitation on the partition. If you want a disk larger than 2 terabytes the only way to achieve that is to either attach a new disk and partition it by using the disk partition tool or remove the existing partition by using the "remove" command and then create a partition.
Note: Removing the existing partition deletes all existing data so any critical data must be backed up to external storage before using this command. For more information on how to take a backup of the data, see To create a backup file of the disk data.
Running the "remove" command asks you for confirmation and once confirmed, it stops all processes (such as MAS subsystems, PostgreSQL processes, and MAS monitor) using the secondary disk. If a swap partition exists and swap is enabled on that, then the swap is disabled.
When you type "y," the command unmounts the disk and removes all partitions on the disk.
The tool restarts the virtual machine after you have removed the partition. You can also manually restart the virtual machine by using the "reboot" command.
To restart the virtual machine
When a partition is created or resized or when swap file is created, the virtual machine needs to be restarted for the changes to take effect. For this purpose, a "reboot" command is provided in the tool.
You are prompted for confirmation and once confirmed, it stops all processes (such as MAS subsystems, PostgreSQL processes, and MAS monitor) and the virtual machine is restarted.
To create a backup file of the disk data
Here are the steps to follow to backup NetScaler MAS data before resizing or removing the partitions.
Note: Creating a backup file requires disk space, and it is recommended that there is enough free disk space available (50% or more) before backup commands are executed.
1. Stop MAS
2. Stop PostgreSQL
su -l mpspostgres /mps/scripts/pgsql/stoppgsql_smart.sh
3. Stop MAS Monitor
4. Create tarball
tar cvfz /var/mps/mps_backup.tgz mps
Note: The operation will take time depending on the size of the data to be backed up.
5. Generate checksum
md5 mps_backup.tgz > mps_backup_checksum
6. Remote copy the tarball and checksum
7. Validate the correctness of the copied tarball. For this generate a checksum of the transferred file and compare with the source checksum
8. Remove the tarball from the MAS VM
rm mps_backup.tgz mps_backup_checksum
In addition to the commands listed earlier, you can also use the following commands in the tool:
To list the supported commands, type help or ? and press enter. To get further help on each of the command press help or ? followed by the command name and press the Enter key.
The "info" command provides information about the attached secondary disk if the disk exists. The command provides the device name, the partition scheme, size in human-readable form, and the number of disk blocks. The scheme can be MBR or GPT. An MBR scheme means the disk was partitioned using an earlier version of NetScaler MAS version. Note that the MBR/GPT based partition can be resized but not beyond 2 terabytes. GPT partition scheme means that the disk was partitioned using NetScaler MAS 12.1 or later.
Note: A GPT partition can be greater than 2 terabytes but when it is created. But you cannot resize the disk to a size greater than 2 terabytes after creating a disk with a smaller size. This is a known limitation of the platform.
The default swap partition on the primary disk of NetScaler MAS is 4 gigabytes and therefore, the default swap space is 4 gigabytes. For the default memory configuration of NetScaler MAS which is 2 gigabytes, this swap space is sufficient. However, when you run NetScaler MAS with a higher memory configuration, you need to have more swap space allocated on the disk.
Note: Swap partition is usually a dedicated partition that is created on a hard disk drive (HDD) during the installation of the operating system. Such a partition is also referred to as a swap space. Swap partition is used for virtual memory that simulates the additional main memory.
Secondary disks that were added in the earlier versions of NetScaler MAS don’t have a swap partition created by default. The "create_swapfile" command is meant for secondary disks created using older NetScaler MAS versions which don’t have a swap partition. The command checks for the following:
- Presence of a secondary disk
- Disk being mounted
- Size of the disk (at least 500 gigabytes)
- The existence of the swap file
The "create_swapfile" command is useful only in cases when the memory is greater or equal to 16 gigabytes and should not be used for cases when memory is low. Therefore, this command also checks for memory before proceeding with swap file creation.
If all the conditions are met, and the user consents to proceed, a 32 gigabytes swap file is created on the secondary disk. Note that the swap file creation process takes a few minutes to complete and take care that you do not interrupt the process while in progress. After successful completion, a restart is done for the swap file to take effect.
After reboot, the increase in swap can be observed using the top command.
To exit from the tool, type exit and press the Enter key.
To attach an additional disk in NetScaler MAS servers in HA
- Shut down each of your standalone nodes.
- In the hypervisor, attach an additional disk of the required disk size to each one of the nodes.
- Now, register your standalone nodes in HA mode.
To create partitions in the additional disks attached to NetScaler MAS in a high availability mode, you must logon to the primary node (using the Floating IP) by using SSH and run the tool there. The changes are automatically synced to the secondary database.