Squeeze - FAQ
What operating systems does Squeeze Support?
Squeeze runs on Windows, Linux or ESXi.
What storage array models does Squeeze support?
Compression estimates are provided for Unity, XtremIO, VMAX and PowerMax arrays. Deduplication estimates are provided for both XtremIO and PowerMax. To obtain the Data Reduction Ratio (DRR) for PowerMax, simply run Squeeze with the –VMAX option.
What is the output of Squeeze?
Squeeze creates a file with the extension .squeeze. This is an encrypted file that when decrypted results in one or more text files. The text file details the estimated results.
How can I decrypt the .squeeze file?
The decryption utility and instructions can be downloaded from the Live Optics page on Inside.Dell.com. The utility is only available to Dell EMC employees.
Can Squeeze scan the whole data on the array from one host or it is required to scan the data provisioned for each host?
Squeeze runs and reports on a single host. If the array is presenting storage to 10 hosts and you want to profile all the data, Squeeze will need to be run on each of the 10 hosts. You would then manually combine the results. Multi-host reporting is a top roadmap item.
If the server has 20 LUNs on the host to scan, where to place the Squeeze executable?
The Squeeze executable can be located on any drive/folder on the host and can be configured to scan all LUNs or specific ones. However if estimating deduplication, the selected drive/folder should have ample free space. Deduplication checking creates a temporary file called the map file which can get large. The larger the data set to scan, the larger the map file will grow.
What is the map file?
When Squeeze is configured to estimate deduplication, it writes deduplication checking data to a temporary file called the map file.
Can the location of the map file be specified?
Yes, by specifying the location in the command line arguments. For example:
squeeze –X2 –dedupe mapfilename - will create the map file in the same location as the executable.
squeeze –X2 –dedupe c:\temp\mapfilename - will create the map file on the C: drive in the temp folder.
This is the same on Linux and ESX. For ESX, we highly recommend placing the map file on one of the datastores as the local disk space for the management console is limited.
Is there a way to force the analysis on a specific volume?
Yes. If you type .\Squeeze.exe –Help at the command or terminal, you will see a command line options.
Squeeze uses 4 type of algorithms (VMAX, unity, x1 & x2). How I can select just one?
You can specify Squeeze to operate in one or more storage platforms with the command line options: -VMAX, -Unity, -X1, -X2. Note that Squeeze will run analysis for all four platforms if you do not specify an option.
Reading the log file, it looks like Squeeze use only the compression algorithm. Is it true or does it also use the X1 or X2 Dedup function?
To run the Dedup function, you must use the option=–DEDUP <~Mapfile Name> where Mapfile is used for the hash table. Please look at the command line options by entering Squeeze –Help for additional details
What data does this “Squeeze” application collect and storage in its results file? Is there a privacy or security risk? What is the potential performance impact on Host/Storage resources ??
The Squeeze tool does not store any customer information nor does it modify any data. It is a read only tool.
When doing compression, the tool reads in blocks of data, runs the compression algorithm and maintains a running record of the compression ratio. Dudupe is similar in that the tool looks at each block and generates a hash code. Squeeze increments a counter if the hash is has been observed previously or creates a new entry if the hash has not been encountered. In both cases, the raw data is discarded once Squeeze has completed its analysis.
The Squeeze output file is also encrypted and only Dell EMC employees are allowed to decrypt the data.
Squeeze does generate a significant IO workload on the volumes that it is reading. This workload is similar in many respects to a backup process. The workload is 100% reads and is generally sequential IO. Additionally, squeeze will use significant CPU resources. In the default setting, Squeeze will fully saturate a single CPU core on a system during its operation.
What file systems are supported by Squeeze?
Squeeze supports both NTFS file systems and ESX 5.0 through 6.0 (ESX 6.5 is not supported). For file systems that are not supported, the user should execute Squeeze with the ignore file system option. Squeeze assumes that the file system zeros out pages that are free and the zero pages are ignored when calculating the data reduction ratio.