StorNext Client System Requirements
To run the StorNext client software, the client system must meet all operating system and hardware requirements.
![Closed](../../Skins/Default/Stylesheets/Images/transparent.gif)
Installing StorNext on all UNIX-based (including Linux) platforms requires installing these quantum-supplied packages:
-
jansson
-
libevent
- protobuf
-
zeromq
These packages are used by StorNext and are bundled with the StorNext binary file.
![Closed](../../Skins/Default/Stylesheets/Images/transparent.gif)
The operating systems, kernel versions, rpm package dependencies, and hardware platforms supported by StorNext File System are listed in the StorNext Compatibility Guide.
Make sure the system meets the operating system and platform requirements outlined in the guide, and, if necessary, update to a supported release or kernel version before installing StorNext.
Note: For Linux clients, you must be disable C-States through kernel boot parameters. See Modify the grub.conf File to Disable the Intel Sleep State.
![Closed](../../Skins/Default/Stylesheets/Images/transparent.gif)
To install and run the StorNext Client software, the system must meet the following minimum hardware requirements.
- For SAN (FC-attached) clients or for LAN clients:
- 1 GB RAM
- 500 MB available hard disk space
- For SAN clients acting as a Gateway server:
- 2 GB RAM
- 500 MB available hard disk space
Note: Gateway servers may require additional RAM depending on the number of file systems, LAN clients, and NICs used. See Gateway Server/Client Network and Memory Tuning for Gateway server/client memory tuning guidelines.
![Closed](../../Skins/Default/Stylesheets/Images/transparent.gif)
RAM, hard disk, and CPU requirements to run the StorNext clients are listed in the StorNext Compatibility Guide.
Note: Make sure the system to be used as a StorNext client meets the hardware requirements outlined in the guide, and, if necessary, update the system before installing StorNext.
For improved performance and stability, on Windows clients it is strongly recommended that CPU C-States be disabled in the BIOS if they are supported by your system. Refer to system vendor documentation for additional information.
![Closed](../../Skins/Default/Stylesheets/Images/transparent.gif)
File System Buffer cache requirements for the StorNext clients are listed in the StorNext Compatibility Guide.
Note: Make sure the system to be used as a StorNext client meets the hardware requirements outlined in the guide, and, if necessary, update the system before installing StorNext.
![Closed](../../Skins/Default/Stylesheets/Images/transparent.gif)
On Linux systems, the StorNext client installation software files are extracted into the /tmp/stornext directory by default. The software, when extracted, requires approximately 60 MB of space, and requires approximately 220 MB of space when installed. Make sure there is enough free space in the default directory to extract the files. If there is not sufficient space, you may need to specify an alternative directory with the -d option.
To specify an alternate location before beginning the installation, enter the following command:
where <archive name
> is the name of the client software file you downloaded, and <dir
> is the directory into which you want to extract the installation files.
![Closed](../../Skins/Default/Stylesheets/Images/transparent.gif)
Additional requirements that must be met for the StorNext system are listed in the StorNext Compatibility Guide.
Note: Make sure the system meets the additional requirements outlined in the guide, and, if necessary, update the system before installing StorNext.
![Closed](../../Skins/Default/Stylesheets/Images/transparent.gif)
Note: The following SNFS mount command settings are explained in greater detail in the mount_cvfs man page.
Multiple mounted file systems typically share a single buffer cache. A StorNext client buffer cache is created for each unique cachebufsize in use. By default, all file system mounts have the same cachebufsize, so they all share the same buffer cache.
The default sizes of the client buffer cache depend on the amount of physical memory detected on your system, as illustrated below.
Amount of Physical Memory Detected | Default Size of the Client Buffer Cache |
---|---|
Less than 2 GB | 256 MB |
Greater than or equal to 2 GB and less than 4 GB | 512 MB |
Greater than or equal to 4 GB and less than 8 GB | 1 GB |
Greater than or equal to 8 GB | 2 GB |
The specific buffer size within the cache defaults to either 256 K (when the total buffer cache size is less than or equal to 1 GB) or 1 MB (for large caches).
Note: Some workflows that do small random I/O could benefit by reducing the cachebufsize to 64 K or smaller, since a small random write within a file requires a read-modify-write of cachebufsize.
In general, increasing the size of the buffer cache will not improve performance for streaming reads and writes. However, a large cache helps greatly in cases of multiple concurrent streams, and where files are being written and subsequently accessed. The total buffer cache size is adjusted with the buffercachecap setting.
The buffer cache I/O size is adjusted using the cachebufsize setting. The default setting is usually optimal; however, sometimes you can improve performance by changing this setting to match the RAID 5 stripe size.
The cachebufsize parameter is a mount option and can be unique for every client that mounts the file system.
Buffer cache read-ahead can be adjusted with the buffercache_readahead setting. When the system detects that a file is being read in its entirety, several buffer cache I/O daemons pre-fetch data from the file in the background for improved performance. The default setting is optimal in most scenarios.
The auto_dma_read_length and auto_dma_write_length settings determine the minimum transfer size where direct DMA I/O is performed instead of using the buffer cache for well-formed I/O. These settings can be useful when performance degradation is observed for small DMA I/O sizes compared to buffer cache.
For example, if buffer cache I/O throughput is 200 MB/sec but 512K DMA I/O size observes only 100MB/sec, it would be useful to determine which DMA I/O size matches the buffer cache performance and adjust auto_dma_read_length and auto_dma_write_length accordingly. The lmdd utility is handy here.
The dircachesize option sets the size of the directory information cache on the client. This cache can dramatically improve the speed of readdir operations by reducing metadata network message traffic between the SNFS client and FSM. Increasing this value improves performance in scenarios where very large directories are not observing the benefit of the client directory cache.
On Windows, to change the size of the pool used by the buffer cache at mount time, modify the Minimum Total Cache Size field on the Advanced Cache Options tab in the StorNext Client Configuration Tool. To change the size of each cache buffer, modify the Individual Buffer Size field on the same tab.
Note: These settings do not apply to Apple Xsan Clients, since they do not use the StorNext buffer cache.