Home
Memory Machine Administration
MemVerge Memory Machine can be installed and run using CLI commands. Many administrative tasks can be performed using the Memory Machine Management Center web interface.
Managing Application Memory
To manage application memory with Memory Machine, you run the application on a data host in an environment with Transparent Memory Service (TMS) and MemVerge Memory Allocation Service (MVMAS).
Configuring Application Memory
Configure managed memory to optimize individual application instances. You can do this by setting the size of the DRAM tier and configuring Linux HugePages. You can make these adjustments "on the fly" while the application instance is running or configure Memory Machine to adust DRAM tier automatically as the application instance runs.
Configuring HugePages
Configure Linux HugePages for MVMAS DRAM caching by calculating how many HugePages your system needs and setting the vm.nr_hugepages kernel parameter to that value.

Overview and Installation
Understanding how persistent memory works is helpful when getting started with MemVerge Memory Machine. You can then install Memory Machine on a server and begin monitoring memory usage and capturing application snapshots.
Using the Memory Machine Management Center
Memory Machine Management Center (M3C) is a web UI application available on the host that is running the Memory Machine Management Service. You can view system resources and performance metrics; create and restore snapshots; schedule jobs; and more in M3C.
Using the Command Line Tools
You can administer and use MemVerge Memory Machine using a Command Line Interface (CLI).
Memory Machine Administration
MemVerge Memory Machine can be installed and run using CLI commands. Many administrative tasks can be performed using the Memory Machine Management Center web interface.
- Upgrading Memory Machine
  To upgrade Memory Machine, run the Memory Machine Package File on each host using the --upgrade option.
- Uninstalling Memory Machine
- Upgrading Application Dependencies
  Upgrading dependency packages for snapshotted applications may render existing snapshots of that application unrestorable.
- Understanding the Memory Machine Database
  Memory Machine Memory Machine Management Center (M3C) uses a database to store information about preferences, statistics, and configurations. You must choose a database system for M3C to use.
- Understanding the Memory Machine Software License
  You must have a valid license to run Memory Machine. The license file is read at startup by the MemVerge Memory Allocation Service (MVMAS). Without a license, MVMAS defaults to running applications using only DRAM.
- Running Memory Machine Services
  Memory Machine data services must be running to enable PMem management, snapshot operations, and connection to the Memory Machine Management Center (M3C).
- Using the MemVerge Memory Allocation Service
  The MemVerge Memory Allocation Service (MVMAS), mvmallocd, is a service that maps programmatic memory to PMem. See oxy_ex-3/topics/a_mvmm_dpme_intro.html.
- Resetting a Host
  Reset a host by removing all snapshot and log data and initializing the DAX devices. This erases the host's PMem.
- Reporting Tiering Statistics
  Use the mvmcli command to view information about paging activity between PMem and the DRAM performance tier for an application.
- Viewing Logs
  Examine the Memory Machine log files to troubleshoot errors – for example, to determine why a license didn't load.
- Managing Application Memory
  To manage application memory with Memory Machine, you run the application on a data host in an environment with Transparent Memory Service (TMS) and MemVerge Memory Allocation Service (MVMAS).
  - Calculating Process Count Limits
  - Running an Application in Memory Machine
    To run an application with Memory Machine memory management, use the mm script supplied with Memory Machine.
  - Stopping an Application Instance
    Stop an application instance in whatever manner is recommended for the application you are running.
  - Configuring Application Memory
    Configure managed memory to optimize individual application instances. You can do this by setting the size of the DRAM tier and configuring Linux HugePages. You can make these adjustments "on the fly" while the application instance is running or configure Memory Machine to adust DRAM tier automatically as the application instance runs.
    - Setting DRAM Tier Memory
      You can enable DRAM tiering and set a maximum tier memory size for each application, enable Memory Machine to apply tiering automatically, or turn tiering off completely.
    - Configuring Automatic DRAM Tier Resizing
      Enable automatic DRAM tier resizing for an application by setting parameters in the application configuration file, mvmalloc.yml.
    - Resizing DRAM Tier Memory
      Resize the DRAM tier memory during runtime to balance performance with memory resource cost using the mvmcli dram resize command.
    - Configuring HugePages
      Configure Linux HugePages for MVMAS DRAM caching by calculating how many HugePages your system needs and setting the vm.nr_hugepages kernel parameter to that value.
- Troubleshooting
- Preparing Intel® Optane™ Persistent Memory
  Prepare PMem for use by configuring the Intel Optane Persistent Memory Direct Access (DAX) devices in AppDirect mode.
- Preparing to Run an Application
  Normally you run applications by using the mm wrapper script supplied with Memory Machine. See oxy_ex-3/topics/t_mvmm_running_an_application_with_mvmm.html. If you do not use the mm to run your application, you must set up the Memory Machine environment manually before running the application.
Working with Applications
MemVerge provides additional instructions for some popular applications supported by Memory Machine, either to assist with complex setup or to support additional functionality.

Configuring HugePages

Configure Linux HugePages for MVMAS DRAM caching by calculating how many HugePages your system needs and setting the vm.nr_hugepages kernel parameter to that value.

About this task

MVMAS uses 2MB translation lookaside buffer (TLB) pages to tier data between PMEM and DRAM. See The Application DRAM Tier. Default page size is 4 kB, which causes excessive page faulting in many memory-intensive applications.

To use HugePages, you must allocate a HugePages pool on the server.

Procedure

Calculate the number of HugePages needed on your system. Since Memory Machine uses 2 MB HugePages, a system requires 500 HugePages for every one GB of DRAM to be used for TLB.
If you plan to use all of the server's memory to run Memory Machine-managed applications, create enough HugePages for 90% of the server's DRAM. (The OS requires the remaining 10% of DRAM to run.) So for example if the system has 64 GB of DRAM, the system requires HugePages for 64 * 0.9 = 57.6 GB, so allocate 57.6 * 500 = 28800 HugePages.
Use the sysctl command to set the vm.nr_hugepages kernel variable to the calculated number of HugePages. For example:
```
# sysctl -w vm.nr_hugepages=28800
```
To automatically set the HugePages at bootup (to avoid having to set HugePages after every restart), add the following line to the appropriate system control file (for example, /etc/sysctl.conf) for your server.
```
vm.nr_hugepages=hugepage_count
```
where hugepage_count is the number of HugePages that you calculated in step 1.