Before you create a replication storage policy you should have already created, at a minimum, a source and target file system. Also, you will not be able to create the policy unless you have specified the blockpool. For more information, see
- If you have not already done so, choose Storage Policies from the Configuration menu.
- Click New.
- Enter the following fields:
- Policy Class: The name of the new policy you are creating
- Policy Type: Click the Replication/Deduplication tab to create a replication storage policy
- The Replication/Deduplication button remains disabled (grayed out) until the blockpool directory has been completely created. Creating the blockpool directory is started on the Storage Destinations page’s Deduplication tab, and proceeds asynchronously as a background job. This background job can take multiple minutes depending on your licensed deduplication capacity. This button will become enabled once that background job completes.
- If there is no deduplication license (for example, if you intend to use replication but not deduplication,) creating the blockpool is still required but the background job will finish within a few seconds.
- File System: Choose from the drop-down list the file system you intend to use as your source
Enter information on the Deduplication, Outbound Replication, Inbound Replication and Source Directories tabs. (See the sections following for more information about fields on these tabs.)
When you are finished entering information about the new policy, click Apply, or click Cancel to exit without saving.
After a message informs you that the policy was created successfully, click OK.
Many fields contain a button that toggles between Inherit and Override. When the button is Inherit, the value for that parameter is "inherited" from the global replication policy that is automatically created when you create a replication file system. To change a parameter's value, click the Inherit button so that it changes to Override, and then enter the new value in the field that appears.
The Deduplication tab enables you to enter parameters for data deduplication and contains the following fields and buttons:
|Deduplication||Click the button beside this field to turn deduplication on or off.|
|Address for Replication and Deduplication||
Enter the address to use for replication to and from this host. In HA configurations this address must match the vIP address. In non-HA configurations the address may be left as localhost.
Note: This field appears only when editing the Replication / Deduplication policy named "Global." To change the virtual IP address from "localhost" you must edit the “Global” policy. Additionally, you must edit the "Global" policy on every file system that has a replication/deduplication policy.
|Minimum File Idle Time Before Deduplication||Specify the interval of time for a file to remain idle before deduplication begins. This field uses the format 00:00:00:00, which refers to days:hours:minutes:seconds.|
|Minimum File Size to Deduplicate||Specify the minimum size a file must be in order to be eligible for deduplication.|
|Filenames Excluded from Deduplication||Specify any filenames you want excluded from the deduplication process.|
|Relative Deduplication Priority||If desired, indicate relative deduplication priority. Preference is given to files from policies with a lower priority number when StorNext considers deduplication candidates.|
|Metadata Content Filtering||When metadata content filtering is enabled, the Deduplication engine will attempt to interpret the content of various backup file formats such as tar and netbackup to extract the component files and apply deduplication algorithms to them. Only use filtering if you are storing this type of file in the policy.|
The Deduplication tab also contains fields which enable you to enter parameters related to data truncation as it applies to deduplication.
Note: If a Deduplication-enabled directory is also a Storage Manager relation point, both StorNext and Storage Manager truncation concepts apply. In this case Storage Manager truncation performs any required truncation, but Storage Manager will not truncate a file unless the StorNext truncation parameters have also been satisfied.
The Truncation portion of the Deduplication tab contains the following fields:
|Truncation||Click the button beside this field to turn truncation on or off.|
|Minimum File Idle Time before Truncation||Specify the interval of time for a file to remain idle before truncation begins.|
|Minimum File Size to Truncate||Specify the minimum size a file must be in order to be eligible for truncation.|
|Files Excluded from Truncation||Specify any filenames you want excluded from the truncation process.|
|Truncation Low Water Mark *||Watermarks help you determine disk space thresholds for your file system. These thresholds determine the point at which StorNext applies or stops applying truncation. At this field, enter the percentage of occupied disk space a file system must reach before StorNext stops applying truncation.|
|Truncation High Water Mark *||Enter the percentage of occupied disk space a file system must reach before StorNext applies or starts truncation.|
|Stub File Size||The stub file is the readable portion of the file that remains after truncation. Enter the desired target size to allocate for the truncated file stub.|
* These fields do not appear when adding a replication storage policy, but are editable when editing the Global replication policy.
The Outbound Replication tab enables you to enter parameters related to outgoing replicated data copied to a destination target and contains the following parameters. For all parameters you can either accept the default values by checking the box to the right of the parameter, or uncheck the box to manually enter the value.
Note: This tab displays replication target machines and files systems that have already been configured on the Storage Destinations > Replication Targets page. You cannot enter replication targets on the Outbound Replication tab; you can only select previously configured replication targets. For more information about creating replication targets, see Configure Data Replication.
|Outbound Replication||Click the button beside this field to turn outbound replication on or off. In this context "On" means that directories associated with this policy will be replicated.|
|Pathname on Target||This is the pathname on the target to which replicated data is copied.|
|Copies to Keep on Target||This is the number of replicated copies created on the target system.|
|Replicate Deduplicated Content||This parameter determines whether deduplicated data is included in the replication process.|
|Use Compression on Network||This parameter determines whether data compression is used prior to and during replication.|
|Use Encryption on Network||This parameter determines whether data encryption is used prior to and during replication.|
|Generate Completion Reports||This parameter determines whether a replication report is generated.|
|Replicate Inline Size||If you specify a size,files smaller than the size you specified will have their contents replicated inline with metadata and will not be transferred separately.|
|'Name' Passed to Target||When entered, a string used to expand the %N parameter in the realization string on the target. StorNext defaults to the source directory path.|
|Filenames Excluded from Replication||Specify any filenames you want excluded from the replication process. This field works the same way as a UNIX shell which lets you pattern match names. For example, entering*o core excludes all .o files and also files named "core." You could also skip all core files by entering rep_skip=core*.|
|Relative Replication Priority||When you enter a priority, preference is given to replicating contents of files from policies with lower priority.|
When deciding whether to create a replication schedule, consider the following:
- This scheduling function schedules the “namespace realization” phase of replication, not replicated data movement. (Namespace realization refers to the process of creating the directory structure for replicated data.)
- If no replication schedule is specified, namespace replications are triggered manually (on demand).
Under Replication Schedules, click New.
Under Replication Targets, select at least one replication target to which the new schedule should be applied.
Enter the parameters for the new replication schedule. Your entries in the Weekday, Day, Hour, and Minute columns determine precisely when replication occurs. You must either select at least one entry in each heading column, or else click All.
To quickly select recurrence, you can click one of the six small boxes under the Use button:
B: Business Days (Monday through Friday every week)
Note: Do not check All under the Minute heading. Doing so can trigger a replication each minute, which is a lot of unnecessary overhead.
(Optional) In the CRON Spec field, input a CRON-like schedule in CRON form by entering an allowable string, and then click Use.
@yearly: Run once a year
@annually: Same as @yearly
@monthly: Run once a month
@weekly: Run once a week
@daily: Run once a day,
@midnight: Same as @daily
@hourly: Run once an hour
After you enter the string, the CRON format will appear at the field, if applicable. For example, if you enter @daily, 0 0 * * * appears.
(Optional) For the Specials heading, check @Reboot to run the schedule once, at startup.
Click Apply to save the new schedule, or click Cancel to exit without saving.
When the confirmation message appears, click Yes to proceed or No to abort.
When a message informs you that the replication schedule was successfully created, click OK to continue.
On the right side of the page is a list of previously configured replication schedules.
- Select the replication schedule you want to modify.
- Click Edit. A page similar to the one on which you added the schedule appears.
- Modify entries under the Weekday, Day, Hour, and Minute columns as desired. (Alternatively, you can enter a CRON-like schedule in CRON format at the CRON Spec field, and then click Use.)
- Click Continue.
- When you are finished making changes, click Apply to save your modifications, or click Cancel to exit without saving.
- When a message informs you that the replication schedule was successfully modified, click OK to continue.
- Select the replication schedule you want to delete.
- Click Delete.
- When the confirmation message appears, click Yes to proceed or No to abort.
- When a message informs you that the replication schedule was successfully deleted, click OK to continue.
The Inbound Replication tab enables you to enter parameters related to incoming replicated data sent from the source blockpool server and contains the following fields:
|Inbound Replication||Turn inbound replication on or off, or disable the process. Quantum recommends not enabling inbound replication when creating new policies, but rather enabling it in the "Target" policy.|
|Pathname on Target||Specify the name of the directory on the target system to which replicated data is saved.|
|Copies to Keep on Target||Specify the number of copies of replicated data to store on the source system. Quantum recommends not specifying this number on the target policy, but specifying it in the source policy.|
Note: See Additional Replication and Deduplication Information
The Source Directories tab enables you to select or create the file system directory used for the policy and contains the following fields and buttons:
|Directory||Select from the pulldown list an existing directory you want to use for the policy.|
|Create Directory Button||To create a new directory, enter a directory location at the field to the left of this button, and then click Create Directory to create the specified directory.|
|Add Button||after either selecting a directory from the pulldown list or creating a new directory, click Add to add the directory as the one used by the storage policy.|
To delete a directory on this tab, select the desired directory from the Source Directories list and then click Delete. Confirm the deletion by answering Yes to the confirmation prompt.
The Blackout tab allows you to schedule times and/or days when you do not want deduplication or replication to run. Specifically, the blackout period specifies periods during which the background data transfers for replication (replication phase 1) do not run. A blackout overrides the replication schedule when there is overlap or conflict. The process for scheduling blackouts is the same for deduplication or replication.
Note: The Blackout tab is not shown when you add a replication storage policy, but it appears when you edit the Global replication policy.
- Select Deduplication Blackout Window and/or Replication Blackout Window to display scheduling options.
- Specify the weekday(s), month(s), day(s), hour(s) and minute(s) when you do not want deduplication and/or replication to run. (For each of these items you can click All.) Alternatively, you can enter a CRON-like schedule in CRON format at the CRON Spec field, and then click Use.
- When you are finished, click Apply to save your changes or Cancel to abort.