Chapter 4 - Configuring the EBR Server
This chapter describes the following system administration tasks for EBR:
You complete most of these system administration tasks using choices from the Server pull-down menu of the EBR Administrator.
EBR requires you to combine clients into backup groups. All clients in a group are scheduled for backup at the same time, although client handling can vary within a group. To start client backups on a staggered schedule, you may assign various clients to different groups.
EBR is shipped with a pre-configured backup group named "Default." You must enable the "Default" group so that the scheduled backup can take place. Scheduled backup of the Default group starts at 3:33 a.m. If this backup start time suits your needs, you do not need to change it.
Important: Initially, the server that hosts the EBR service is assigned to the default group as a client of the EBR service it hosts. You may change this if you wish, but make sure your EBR server belongs to a group that is enabled for a scheduled backup. Otherwise, on-line indexes will not be backed up.
To ensure that an EBR server is enabled for backup as part of the default group, follow these steps:
1. Choose Setup from the Client menu. The Client Dialog Box appears.
2. Inside the Clients dialog box, highlight your server name and click the Edit button. The Edit Client Dialog box appears.
3. Be sure the Default box is checked inside the Group field.
4. Choose Groups from the Client menu. The Groups dialog Box appears.
5. Highlight the Default group and click the Edit button. The Edit Group dialog box appears.
6. Change the Autostart field from Disabled to Enabled.
The Start time field controls the beginning time of a backup group.
If you want to segregate the backup volumes for specific groups, or if you prefer to start backing up certain machines at different times, you can do so. For example, you may want to start the backup of the Administration department at noon, and change the backup volume before you go home. Then you could schedule backup of the Engineering department to start at midnight.
The Client retries field controls the number of times EBR will try again in case of an unsuccessful backup. By default this number is zero. Setting this number higher might result in fewer unsuccessful backups, but will probably slow down completion of the backup group.
After configuring backup groups, you assign a client to a backup group using the Clients dialog box. This procedure is described in the section "Configuring the Server' s Clients" later in this chapter.
Tip: Every client must be in an enabled group to be part of scheduled backups. Use the Group Control command to see the clients in each group, and whether they are being backed up.
An EBR service can have any number of backup groups. The service will start backing up all clients in the group at the group start time. You may start multiple backup groups at the same time. This practice is not recommended, however, because the group backup will take more time to complete than if you had started each backup at a different time. Each group completes before the next one starts.
To create a backup group and set the time for that group' s backup start time, follow these steps:
1. Choose Groups from the Client menu. The Groups dialog box appears, displaying the names of all the backup groups known to the EBR service:
2. Click the Create button to form a new group. The Create Group dialog box appears:
3. Use the Name field to enter a name for the backup group.
4. Use the Autostart choices to enable or disable backup, as follows:
Enabled - allows the network-wide backup to start at its scheduled time.
Disabled - prevents the scheduled backup from starting.
5. Use the Start time field to enter the time you want the EBR service to start backing up this group. The default is 3:33 a.m.
6. Enter a group name into the Name field. Choose a name that helps you keep track of group membership.
7. Click the mouse or Tab to highlight the Autostart field.
8. Press the [e] key to enable backups, or use the mouse to select Enabled from the drop-down menu. (You could press the [d] key or select Disabled with the mouse to disable backups.)
9. Enter a new time into the Start Time field. Use the 24-hour clock convention to assign the new time. For example, type "13:00" for 1:00 p.m. or "1:00" for 1:00 a.m.
Tip: To help reduce network traffic, it is best not to have multiple backup groups starting at the same time.
10. Click OK or press ENTER to apply the change.
Note: If a message box appears saying "user <user_name> on machine <system_name> needs to be on administrator list," you do not have permission to make configuration changes. See "Adding or Changing Administrators" later in this chapter for more information.
Your new group name appears in the Groups dialog box.
After you are finished creating and setting the start times for the backup groups on your network, you may assign clients to the groups. See "Configuring the Server's Clients" later in this chapter for information on assigning clients to backup groups.
Changing the Backup Time for a Group
To change the start time for a group' s network-wide backup, follow these steps:
1. Choose Groups from the Client menu.
2. Select the group name for which you want to change the start time.
3. Click Edit or press the [Enter] key to bring up the Edit Group dialog box:
4. Use the [Tab] key to move to the Start time field.
5. Enter a new time into the Start Time field. Use the 24-hour clock convention to assign the time.
6. If you want to retry after an unsuccessful client backup, set Client retries to one or more. (However, this activity might slow down group backup.)
7. Click OK or press the [Enter] key to apply the change.
Important: If you fail to list a time in the Start time field, the group backup will not take place.
You cannot delete a group if any clients are assigned to that group, or if the group is named "Default." To see if any clients belong to the group you want to delete, look at the Group Control dialog box. Before deleting a group, you can always move clients to a different group using the Edit Client dialog box.
To delete a backup group, follow these steps:
1. Choose Groups from the Client menu.
2. Select the name of the group you want to delete.
3. Click Delete to delete the group. If any clients belong to the group you are trying to delete, an "unable to delete" message appears.
Setting up the Backup Schedules
EBR automatically backs up all the clients in a group, starting at the time you specify within the Groups dialog box. The type of backup (full, incremental, or level) is determined by the individual backup schedules for the clients.
This section provides detailed information on the components of the Schedules dialog box and the logic behind setting up a schedule. If you prefer, you can skip ahead to the section "Creating a New Backup Schedule" later in this chapter for a summary of how to create a schedule.
EBR is shipped with several schedules already set up for you to use as is or as templates for making your own schedule.
To see the default schedules, complete the following steps:
1. Choose Schedules from the Server menu. The Schedules dialog box appears:
This dialog box displays the names of the schedules identified by the EBR service you are administering. Five backup schedules are provided for your convenience. You may delete all of the schedules except the one named "Default."
Before you create a new schedule, look at the pre-configured schedule named "Default."
2. If it is not already highlighted, highlight Default, then click Edit or press the [Enter] key. The Edit Schedule dialog box appears, displaying a calendar of the current month:
The Name field displays the currently selected schedule name.
Use the Period field to set the calendar to a weekly or monthly cycle. A period is the duration over which backup cycles are repeated. If you select Week, the schedule you set repeats every week. If you select Month, the schedule you set repeats every month.
Each calendar day displays a letter next to the date. An "f" means full backup, an "i" means incremental backup, and an "s" means "skip the backup." In addition, there are nine other backup levels, each represented by a number from 1 to 9. See "Understanding Backup Levels" later in this chapter for more information about backup levels.
An asterisk (*) next to a backup level means you have forced a different level (an override) on that day. For example, if you schedule a full backup on every Monday of the month, but the second Monday for a particular month is a holiday, you can force a "skip backup" for the second Monday, and force a full backup on Tuesday instead.
3. Click the Prev Month and Next Month buttons to move back and forth one month at a time.
4. Click Cancel to close the Edit Schedule dialog box, or click OK to save your editing changes.
Logic Behind the Scheduling Schemes
It is very easy to set up a schedule using EBR Administrator. But first, you need to decide on a schedule that fits your backup requirements. When configuring backup schedules, you must consider:
![]()
How long do you want to keep backed-up data? ![]()
How many versions of the data do you want to maintain? ![]()
How much data do you have to back up? ![]()
How many backup volumes do you want to use? ![]()
How long does it take to complete network-wide backups? ![]()
Do you want to be able to recover from a disk crash using relatively few backup volumes?
For example, the typical capacity of an 8mm tape is approximately 5 gigabytes, and the maximum transfer rate is about 500 kilobytes per second. Systems generally cannot sustain that transfer rate, but even if they could, it would take more than 5 hours to back up 10 gigabytes of data. To back up a network with a lot of data, you may assign clients to different schedules, and stagger full backups over several days.
Determine how many backup volumes you want to keep - this number depends upon how often the data changes, and how long you want to keep the backups on-line.
If you schedule a full backup once a month followed by incremental backups every night, you will need more tapes to fully recover from a disk failure. Suppose a site has 10 gigabytes of data, and 5% of all the data is modified each day. Thus, 500 megabytes of data need to be backed up every day. This amount of data fills up about one eighth of a regular 8mm tape. If you are retaining media for three months, you will have about 25 backup to keep on your shelves.
You also should decide on a policy for recovering files. For example, if users expect to be able to recover any version of a lost file for at least three months, you must maintain all the backup volumes for the three month period.
If you want to maintain fewer backup volumes on your shelves, you can use level [1-9] backups between the full and incremental backups. However, you must use the full backup volume and all the incremental backup volumes to recover from a disk crash.
EBR supports four kinds of backup levels:
![]()
Full - Every file is backed up, no matter when it last changed. ![]()
Level [1-9] - Files are backed up if they have changed since the last backup at a lower numeric level. For example, if you take a level 1 backup, then a level 9, then a level 8, the level 8 backs up all files changed since the level 1 backup. Therefore, lower levels represent fuller backups, and include more data. ![]()
Incremental - Files are backed up that have changed since the last backup, regardless of level. ![]()
Skip - No files are backed up; the backup is skipped. You may want to skip a backup on a holiday if you know that no one will be around to change backup volumes on the server.
The diagram that follows illustrates how backup levels work. You may use a backup scheme that includes occasional full backups followed by level [1 to 9] and incremental backups during the cycle. Different backup levels allow you to trade off the number of backup volumes and amount of time required to complete a backup, versus the number of backup volumes and amount of time it takes to recover from a loss of files.
In this example, assume you use a new tape for each day's backup. However, using a new tape for each day's backup is not required. On day 1, a full backup runs. On day 2, the incremental backs up everything that has changed since the full backup. On day 3, the incremental backs up everything that has changed since day 2. On day 4, the incremental backs up everything that has changed since day 3. At this point, you have four tapes. To recover from a crash, you need all four tapes - the one with the full backup (day 1), and all the tapes with incremental backups after that.
On day 5, the level 9 backs up everything that has changed since the full backup. You no longer need data on the tapes from days 2, 3, or 4. To do a full recovery, all you need is the full backup tape and the level 9 backup tape. If you had to recover from a complete loss of a disk, you have reduced the number of tapes you need to two.
On day 9, the level 8 backs up everything that has changed since the full backup. Again, you only need two tapes to recover a disk: the full backup, and the level 8. A similar sequence of events occurs on day 13 with the level 7 backup.
Level [1 to 9] backups help you maintain control of the backup volumes in your media set. If you need to recover everything on a disk, you should be able to do so with four sets of backup volumes or less, depending on how carefully you plan your backup strategy.
A full backup is theoretically equivalent to a level 0 backup. Note that in the example, three level 9 backups have the same effect as the level 9, 8, and 7 backups. Each level 9 saves files changed since the last backup of a lower numeric level, which in this case is the full (level 0) backup.
EBR provides a backup schedule named "Default" that you can change, but not delete.
The default schedule repeats on a weekly basis. Every week, a full backup takes place on Sunday, followed by an incremental backup every remaining day of the week. This schedule is useful for small-to-medium sized networks.
In addition to the default schedule, EBR is shipped with four other backup schedules you can use as they are, or modify as required:
![]()
Full Every Friday is similar to the default, except that full backups are on Friday instead of Sunday. ![]()
Full on 1st of Month does a full backup once a month, and incremental backups all other days. ![]()
Full on 1st Friday of Month is similar, except it starts on Friday instead of the first of the month. ![]()
Quarterly does a full backup once a quarter, level 5 backups monthly, level 7 backups weekly, and incremental backups every other day.
If you have a large network with multiple EBR services and want to stagger the full backups, you may do so by using multiple backup schedules. For example, you could assign some clients to a "Full on Wednesday" schedule and other clients to a "Full on Friday" schedule.
Creating a New Backup Schedule
This section illustrates setting up schedules with two examples. Read through the examples before you try to set up any new schedules on your own. There are four basic steps to set up a schedule:
1. Pick a name for the schedule.
2. Decide on a period.
3. Select backup levels for each day in the period.
4. Set the overrides, if any.
Note that when you set up one month in the Schedules dialog box, that setup applies to all calendar months, and carries over from year to year. In the same way, if you set up a weekly schedule, it applies for all calendar weeks. This strategy allows you to plan all the weeks or months in one session, without having to repeat the process every week or month.
Important: You can make changes to an existing schedule, but you may not change its name. If you want to change the name of a schedule, you must delete it and re-create it with a new name.
This schedule is suitable for EBR clients with files that change often. Suppose you want to schedule a full backup every Monday because weekly full backups make disaster recovery easier than monthly full backups. In addition, someone is available to change backup volumes and monitor the EBR service on Mondays. Other days of the week, incremental backups protect the daily changes. You are going to name the backup schedule "Full on Mondays" to remind yourself that this schedule calls for a full backup every week on Mondays.
To create a weekly schedule, follow these steps:
1. Select Schedules from the Server menu.
2. Click Create to devise a new schedule. The Create Schedule dialog box appears:
3. Enter "Full on Mondays" into the Name field.
4. In the Period field, select Week if it is not already selected.
In the calendar section, the letter "i" next to each day represents an incremental backup for that day. Each Sunday of the month has a letter "f" next to it, representing a full backup on Sundays. Exchange the "f" and "i" between Sundays and Mondays.
5. Select any Sunday in the month.
6. Click the Set Level button. The Set Level dialog box appears:
7. Click the Incremental button. Then click OK or press [Enter]. All the Sundays change to "i." When you create a weekly schedule, a change made to one week is also made to all the weeks in the calendar.
8. Select any Monday in the month.
9. Click the Set Level button again. The Set Level dialog box appears.
10. Click the Full button, then click OK or press [Enter]. All Mondays change to "f" to reflect full backup.
Scroll through all the months, using [P] for Previous month and [N] for Next month, and notice how the schedule you set up for the current month is maintained throughout the entire year and for years to come.
11. In the Create Schedule dialog box, click OK or press the [Enter] key to save the schedule.
Note: If the message "user <user_name> on machine <system_name> needs to be on administrator list" appears, you do not have permission to make configuration changes. See "Adding or Changing Administrators" later in this chapter for more information.
This monthly schedule is for the clients with files that do not change very much, and therefore need a full backup only once a month. A full backup takes place on the first day of the month. In the middle of the month, a level 5 backup backs up all the files that have changed since the full backup. On other days of the month, an incremental backup protects daily changes.
You may want to consider naming this schedule "Monthly Fulls" to remind you that it calls for a full backup only once a month. To demonstrate how overrides work, you will override the backup scheduled on the sixteenth day. Overriding the backup schedule lets you have more control over the backup schedule. This approach lets you skip backups on company holidays. Follow these steps to override the backup schedule:
1. Select Schedules from the Server menu.
2. Click Create to create a new schedule.
3. In the Create Schedule dialog box, enter "Monthly Fulls" into the Name field.
4. In the Period field, select Month.
Notice that the month is already set up to do a full backup on the first day.
5. Set up the month for a level 5 backup on the fifteenth day.
6. Use the mouse, or press [Tab] and then the down and right/left arrow keys, to highlight the fifteenth day of the month (15/i).
7. Click the Set Level button to bring up the Set Level dialog box.
8. Click the button labeled 5, or type [Alt]+[5] to change to a level 5 incremental backup. Then click OK or press [Enter].
Using the Prev Month and Next Month buttons, check the schedules for past and future months. Notice how the monthly schedule is retained every month - a full backup on the first day, a level 5 on the fifteenth day, and incremental backups every remaining day of the month.
Return to the current month and override the scheduled backup on the sixteenth day of the month by skipping a level to bypass the backup for that day. Follow these steps to bypass a level:
1. Select the sixteenth day of the month (16/i).
2. Click the Set Level button to open the Set Level dialog box.
3. Click the Skip button to override the incremental backup with a "skip the backup" instruction for that day.
4. Click the Set Override button to close the Set Level dialog box and return to Create Schedule Window.
The calendar date should look like this: 16/s*
Using the Prev Month and Next Month buttons, check the schedules for past and future months. Notice that the change you made is not carried through the schedules for these months. An override is a special case - it is a way to customize the calendar for special events or circumstances.
5. Click OK or press [Enter] to save the schedule you just created.
Important: Remember, any override must be scheduled explicitly each time. In other words, an override does not automatically repeat in a monthly or weekly period.
You may set up as many EBR backup schedules as you like, as long as you give each one a unique name. To assign a client to a backup schedule, you must go to the Clients dialog box. See "Configuring the Server' s Clients" later in this chapter for more information.
To create a backup schedule, follow these steps:
1. Choose Schedules from the Server menu.
2. Click Create to create a schedule or Edit to modify a schedule.
3. Assign the schedule an appropriate name in the Name field of the Create Schedule or Edit Schedule dialog box.
4. Select a weekly or monthly period for the schedule.
5. Select a date within the calendar and set its backup level using the Set Level dialog box.
- Select Full for a full backup on a calendar day.
- Select Incremental for an incremental backup that day.
- Select Skip to avoid a backup on that day.
- Select a number between [1] and [9] for a particular level.
6. After selecting a level, click OK to set the backup level. Or select a level, then click Set Override to override for that day.
7. To remove an existing override, click Clear Override.
8. In the Create Schedule or Edit Schedule dialog box, click OK or press the [Enter] key to save your custom schedule.
EBR Administrator lists the devices used for backups and recovers in the Devices window. If you change a device on the backup server, you must register the change with the EBR service before bringing the server back on line.
EBR supports all tape devices certified for use on VINES 6.0 and later releases. This includes QIC, 4mm and 8mm digital tape.
To view devices recognized by the EBR service, take these steps:
1. Choose Devices from the Server menu. The Devices dialog box appears:
This dialog box displays all backup devices the EBR server recognizes, for backing up and recovering client files.
2. If there are multiple devices, highlight the one for which you want more information.
3. Click Edit or press the [Enter] key. The Edit dialog box appears for devices:
The Name field displays the device type; the name cannot be changed.
The Media type field displays the type of backup media used by this device. The choices represent quarter-inch cartridge, four millimeter tape, eight millimeter tape, high capacity 8mm, and optical disc, respectively. The Media type cannot be changed.
Click Field Help if you need more information about either field.
Contact your Banyan authorized reseller for the list of currently supported backup devices, or see the EBR Release Notice.
Note: The Media type defaults to 8mm tape. If this device is incorrect, use the Devices window's Delete button to delete the device, then choose Create to create a new device with the correct Media type.
Parallelism is an EBR service feature that backs up multiple save sets (or clients) in parallel and multiplex files onto the same backup volume. This feature keeps a steady stream of files supplied to the backup device so that it can operate at higher speeds. In addition, this process ensures that no particular client can monopolize the server. This means other clients can back up files to the server simultaneously. EBR is shipped with a default parallelism equal to 2, in other words EBR can back up a maximum of two clients at a time.
To optimize the efficiency of the server, you can control how many save sets can be backed up at one time. You may want to adjust the number of parallel save sets if:
![]()
The server takes too long to back up all its clients (increase the parallelism) ![]()
The server is so busy backing up clients it is useless for any other tasks (decrease the parallelism)
To optimize the efficiency of the server, you can change the value of the Parallelism field in the Server window. To change the number of save sets that can simultaneously be sent to the EBR server, follow these steps:
1. Choose Setup from the Server menu. The Edit Server dialog box appears.
2. Select the Parallelism field.
3. Increase or decrease the number depending on whether you want to increase or decrease the parallelism. Parallelism may be any number from 1 to 5.
4. Click OK or press the [Enter] key to save your change.
Note: A high value for Parallelism allows savegroups with many clients to complete in less time, but consumes more server resources (for example, swap space, communications buffers, and so on) during the backup.
Adding or Changing Administrators
When you first create an EBR service (for example, on the server named Redwood@Servers), only members of the server' s AdminList (AdminList@Redwood@Servers) are authorized to change the EBR service configuration, or to use EBR to back up and recover data from Redwood@Servers. AdminList members can add, delete, or change permissions to give others EBR administration privileges. To change permission settings, follow these steps:
1. Choose Setup from the Server menu.
2. Highlight the Administrator field, then scroll to the end of the list.
The Administrator field displays the names of the users or user groups who have permission to change the configuration of the EBR server.
3. Enter a new user name into the Administrator field. This name can be a user's StreetTalk name or a StreetTalk list. It can also be root or root@host if you want this EBR service to interoperate with Legato® NetWorker® running on the UNIX system named by host.
4. Click OK to save the change. Every user listed in the Administrator field now has permission to modify the EBR service configuration.
Note: Adding a user or list name to the an EBR service's Administrator list only gives these users or list members administration rights. It does not allow them to back up or recover server data. This right is initially limited to members of the server's AdminList.
Automatic Index Management Policies
This section describes how to create policies that automatically manage the on-line indexes for a client. You can also manage the indexes manually using the Indexes and Volumes dialog boxes. Manual index management as well as additional concepts of index management are described in Chapter 5.
Whenever an EBR service starts a backup, it creates entries for the saved files in the on-line indexes. The service maintains two types of on-line indexes: a file index and a media index. The file index stores information about files backed up by the EBR service, and the media index maps the saved files to backup volumes. The service maintains one file index per client, and one media index per EBR service. The service uses the on-line indexes as databases to locate files for recovery.
Every backup adds entries to the indexes, so the indexes require management to control their growth. The EBR service automatically manages the contents of the indexes with policies that you specify for each client. Automatic index management policies preserve valuable disk space by removing expired entries from the indexes. You may use EBR Administrator' s pre-configured policies or customize your own policies using the Create Policy dialog box.
Each client requires two separate policies: a Browse policy and a Retention policy. The Browse policy determines how log entries for the clients' files remain in the on-line file index so they can be accessed by EBR or the server console recover browser. Entries younger than the Browse policy are called browsable. Entries older than the Browse policy are automatically removed from the on-line file index. Entries older than the Browse policy but not older than the Retention policy are called retrievable. The Retention policy determines how long entries can be retained in the on-line media index and thus be retrievable. Entries older than the Retention policy are marked as recyclable in the media index.
When all entries on a volume become recyclable, the backup volume becomes "recyclable" and can either be stored in a safe place, or rotated. To rotate the volume, re-label the volume and re-use it or both. We recommend that you store tapes, because they are inexpensive, can hold a large capacity of data, and may be useful in the future. If you are concerned about tape costs, you may consider rotating the backup volumes in your media set. Use the Volumes dialog box to determine which volumes are eligible for recycling. In addition, you can use the Volumes dialog box to remove their entries from the on-line indexes.
The following diagram illustrates the interaction between an entry in the indexes and the index policies.
After you create the policies for automatic index management using the Create Policy dialog box, the policies appear as selections for both the Browse and Retention policies in the Edit Client dialog box. Every client requires both a Browse and Retention policy. When you select browse and retention policies with the Clients command, you are telling the EBR service the minimum time period for entries to remain browsable and recoverable for that client.
During the normal index checking that occurs when the EBR service starts up and when a scheduled backup completes, the EBR service compares the backup date of entries in the indexes with the time period specified by these policies. If an entry has passed the time period specified by either policy, it is automatically removed from the appropriate index, thus freeing up space on the disk.
To create policies for automatic index management, choose Policies from the Server menu. The Policies dialog box appears:
The Policies dialog box displays the names of all policies known to the current EBR service. If you have not created any custom policies, the dialog box displays these five pre-configured policies: Decade, Month, Quarter, Week, Year.
To see the details of a policy, highlight its name in the Policies dialog box and click Edit or press [Enter]. The Edit Policy dialog box displays the name of the currently selected policy, its period, and number of periods. To close the dialog box, click Cancel.
To create a new policy, follow these steps:
1. Choose Policies from the Server menu.
2. Click Create. The Create Policy dialog box appears:
3. Enter a new name for the policy in the Name field.
4. Select a Period for the new policy. Your choices are Days, Weeks, Months, and Years.
5. Enter the Number of periods for the policy by entering a number in that field. For example, 26 Weeks produces a policy that could be called "Semiannual" (for half a year).
6. Click OK or press the [Enter] key to save your work. The newly created policy appears in the Policies dialog box and in the Clients dialog box as a Browse policy and a Retention policy.
Enter any name for the policy that you want. However, you should describe the time period to help you remember the length of the policy. For example, when you specify Quarter as a Browse policy for a client, it is easy to tell that the entries are kept in the on-line file index for three months.
To delete a policy, follow these steps:
1. Highlight the undesired policy in the Policies dialog box.
2. Click the Delete button. EBR Administrator deletes the policy.
Tip: You are not allowed to delete policies currently in use, or the default policies. The default policies are Quarter for Browse policy and Year for Retention policy.
To change a policy, follow these steps:
1. Highlight the desired policy in the Policies dialog box.
2. Click Edit or press the [Enter] key. The Edit Policy dialog box appears:
3. Select the Period and Number of periods for the policy. Its name cannot be changed.
4. Click OK or press the [Enter] key when you are done.
Note: An EBR service's Retention policy must exceed the Browse policy of all its clients. Under automatic index management, every related file index entry must be purged before backup media can be recycled.
Selecting Policies for a Client
After you have created policies using the Create Policy dialog box, select them for a client as a Browse policy or Retention policy.
To select policies for a client, choose Setup from the Client menu, highlight a client name in the Clients dialog box, and click Edit or press the [Enter] key to bring up the Edit Client dialog box.
Notice the Browse policy and Retention policy selections in the Edit Client dialog box. The five pre-configured policies and the custom policies you created appear as choices for both Browse policy and Retention policy selections.
You may select a different Browse and Retention policy. The default selections are Quarter for Browse policy and Year for Retention policy.
Important: A client's Browse policy may not exceed its Retention policy. An error message appears if you attempt to assign a Browse policy to a client longer than the Retention policy assigned to that client. For example, one month for the Browse policy and one week for the Retention policy is unacceptable because you cannot assign browsing for a month on tapes only saved for a week.
Using Global Backup Directives
An EBR service's Global Backup Directives contain instructions on how to back up specific files and directories. They're called Global Backup Directives because they apply to all of a service's clients. There are also local backup directives, which override the global ones and apply only to a specific server or workstation Workstation clients may set up backup instructions using the Local Directives command in EBR. Local directives for VINES servers are preconfigured based on the services they host. These directives normally do not need to be altered.
Tip: Effective use of directives requires extensive knowledge of a client's file system layout, including the names of files, what they do, and so on. Many system administrators know workstation file systems. Few administrators, however, are familiar with the internals of VINES servers. EBR is shipped with default global and local directives that have been tailored to work well on all servers. It is not necessary to alter them in any way. The information in the rest of this section is primarily intended for administrators who either want to customize DOS Standard Directives or have the appropriate level of knowledge of VINES server file systems to allow them to modify the VINES directives without adversely affecting backup performance and integrity.
Directives are very flexible. You can apply them to specific directories or to disk volumes. You can use pattern matching to apply instructions to specific files in the directory tree. For example, using "skip: *.tmp" skips the backup of any file with suffix ".tmp" in the directory to which this directive has been assigned.
Note: This type of pattern matching to exclude files from a scheduled nightly backup may only be applied to all clients of an EBR server, not to individual clients. They are global instructions that apply to clients for which you have configured directives on the EBR server.
EBR is shipped with a set of pre-configured directives. Use the Default and Default with compression directives for a VINES server, and the DOS standard directive for DOS and Windows workstations. If the service is running on a server with TCP/IP support, specify that the Unix and Unix with compression directives for UNIX systems hosting NetWorker clients, and the NetWare standard directives for NetWare servers hosting NetWorker for NetWare are accessible over TCP/IP.
To open the Directives dialog box, follow these steps:
1. Choose Directives from the Server menu. The Directives dialog box displays the types of global directives recognized by the EBR service.
2. Select one and click Edit or press the [Enter] key to open its Edit Directive dialog box. The Edit Directive dialog box appears:
The Edit Directive dialog box shows you the contents of the currently selected directive. Each directive in the list is preceded by the directory to which it applies (for example, <</>>).
3. Use the arrow keys or buttons to scroll through the contents of the Directive field.
Note: Any directive that you enter under "<</>>" (slash) applies to all disk volumes for the client.
You can type directly into the Directive field - it is a text window. The syntax for directives is as follows:
1. Enter the volume or directory name to which you want the directive or set of directives to apply. For example:
<<"C:">>
2. Enter the name of the directive under the directory name, and include the pattern-matching instructions, if any. For example:
skip: *.exe
This instruction means to skip all the files with a suffix of .exe.
There are many directives that EBR uses internally to help it optimize server backups. Only two are likely to be useful for most administrators. They are:
skip skips the backup of a named directory or file
always always backs up the named directory or file, no matter what backup level is scheduled.
Prepending a "+" to either of these directives means that the directive applies in the specified parent directory and all of its child directories. For example, the directive:
<</disk1>>
+skip: core
means that any file or directory under /disk1 named core will not be backed up.
The other directives you see in the dialog box are for EBR internal use. You should not change or delete them.
The EBR service can be configured to notify selected users of various noteworthy events in a variety of ways. By default, notifications are turned off in EBR. You may enable any or all of EBR's notifications using the Notifications dialog box. To open the Notifications dialog box, choose Notifications from the Server menu of EBR Administrator.
The Notifications dialog box lists all the types of notifications supported by the EBR service:
Each of the lines listed in the Notifications dialog describes a condition that the EBR service can notify you about with a message describing the condition. These conditions are:
Index Size: Index size notifications are sent when disk space for the EBR service is getting low or there is no more disk space for its on-line indexes.
Log Default: A condition that EBR services associate with a variety of status messages of the type normally placed in service logs. All types of notifications are logged, even if they are also delivered by some other means.
Savegroup completion: The EBR service sends this notification when a scheduled savegroup completes. The notification contains the completion status of the savegroup, including descriptions of any errors that occurred.
Tape mount request 1: These notifications describe ongoing tape operations, including conditions that may require attention such as "soft" media errors or tape mount requests for recovery.
Tape mount request 2: These notifications describe more pressing tape requirements, including mount requests originally posted to Tape mount request 1 that are still outstanding after 15 minutes.
Tape mount request 3: These notifications describe critical tape mount requirements, including mount requests for backups and mount requests for recovers originally posted to Tape mount request 1 that are still outstanding after 30 minutes
To enable any of these notification types, or to change the method of notification, select the notification you want to enable or modify, then choose the Edit button in the Notifications dialog box. The Edit Notifications dialog appears.
Each notification has a Name and associated Action.
You cannot change the name of any of the default notifications, but you can change the action to any of the following:
none No action. This is the default for all notification types when an EBR service is first created
ebrmail Mails the notification text you specify to the user or list you specify. The syntax for this action is:
ebrmail [-s subject ] recipient
where subject is the text you want to appear in the subject field of the message and recipient is the StreetTalk name of a user or list of users who should receive the notification. If subject contains multiple words, they must be delimited by quotes. For example:
ebrmail -s "EBR mount request" AdminList@Redwood@Servers
is an appropriate action for a tape mount request.
ebrmesg Sends the notification as a 25th-line or Windows popup message. The syntax for this action is:
ebrmesg recipient
where recipient is the StreetTalk name of a user or list of users who should receive the notification
ebrlog Logs the notification to the service' s logfile
Note: If you specify the action for a notification with incorrect syntax, the notification will fail and an error will be logged in the EBR service' s logfile. When the EBR Mail User cannot log in, notifications that have been configured to be mailed will, instead, generate an ebrmesg reminding you to configure the EBR Mail User.
Configuring the Server' s Clients
Before a workstation client or VINES server agent can be backed up by an EBR service, an administrator must determine the following:
![]()
Agreed-upon name for each workstation client ![]()
Name of the EBR service that will back it up ![]()
Backup schedule ![]()
Browse policy and retention policy ![]()
Set of directives ![]()
Backup group ![]()
Save set (if not All) ![]()
What other clients can recover this client's data
There are predetermined values for all these items except the client name. When you install EBR on a server, it automatically creates a client entry for itself. In other words, the server itself is added to its list of clients, and its disk volumes will be backed up by EBR according to a schedule named "Default." However, you must add all other clients to the EBR server using the Create Client dialog box.
Before a workstation client can be backed up automatically by the EBR service, it must be running Scheduled Backup.
The client configuration tasks are completed in the Edit Client or Create Client dialog boxes.
To see the Edit Client dialog box, complete these steps:
1. Select Setup from the Client menu.
2. In the Clients dialog box, click the Edit button. The Edit Client dialog box appears:
In the Create Client dialog box, the Name field is empty. You must enter the name of the client into this field. (You and the workstation user must agree upon the name.)
The Server field displays the name of the currently selected EBR server.
The Schedule field displays all schedules known to the EBR service. EBR is shipped with five pre-configured schedules for your convenience. Any other schedule you create using the Schedules dialog box also appears here as a choice.
The Browse policy and Retention policy fields display the policies for automatic index management recognized by the EBR server. EBR is shipped with five pre-configured policies for your convenience. Any other policy you create using the Policies dialog box also appears here as a choice.
The Directive field displays the directives recognized by the EBR service. There are four pre-configured directives shipped with EBR. Any new directives you create using the Directives dialog box also appear here.
The Group check boxes display all the backup groups known to the EBR server. These groups are created in the Groups window.
Note: A client can belong to more than one group, if you want to split the backup of the client' s file systems over different schedules. See "Scheduling Large Server File Systems" later in this chapter for details.
The Save set list displays the file systems that the EBR service backs up for the client. The default "All" means "back up all file systems for this client." You can change this default by typing over the word "All" in that field. For clients that are DOS or Windows workstations, you may replace All with any valid DOS path on the client (C:\IS_FILES, for example).
The Recover access list (optional) displays the names of other clients that are allowed to recover this client' s data (an operation known as a directed recover). This item is only meaningful when the client names you add are those of VINES server agents, because any administrator of an EBR service can initiate a directed workstation recover using the workstation EBR program. A workstation cannot recover data saved from a server, even if you give it the necessary permission.
To configure a new client, all you need to do is provide the client' s name and make choices in the Create Client dialog box.
To modify an existing client, select its name in the Clients dialog box, open the Edit Client dialog box, and edit the client's name.
Using StreetTalk Attributes to Change Workstation Client Defaults
EBR workstation clients look for various StreetTalk attributes when they start up. These clients use the value of these attributes to determine certain configuration parameters. These parameters do not apply to EBR agents. You do not need to set any of these attributes if the default configuration is acceptable. You use the MATTR command to set StreetTalk attributes. The following attributes are recognized by EBR workstation clients:
attribute = value | Attribute of | What it means |
<22:0> = name | Group | Specifies the name of the EBR service that backs up workstation clients owned by members of this Group. |
<22:0> = 48 | EBR service | Disables all workstation backup and recover access to this EBR service. Does not disable access by EBR Administrator. |
<22:0> = 49 | EBR service | Specifies that workstation clients of this EBR service can only back up and recover data on their local hard disk drives (the default) |
<22:0> = 50 | EBR service | Specifies that workstation clients of this EBR service can back up data from local hard disk drives and network drives. |
Note: An EBR service also uses service attributes 22:1 to 22:200 to record the names of its workstation clients. You should not use any of these attributes numbers for any other purpose, nor should you attempt to edit them manually. If any of these service attributes is corrupted, the EBR server no longer recognizes workstation clients whose identities were recorded in those attributes, and prompts the client for a name. If this happens, supply the name the client has been using, and the service re-initializes the attribute data.
To delete a client from the EBR service's client list, follow these steps:
1. Select Setup from the Client menu.
2. Highlight the client to delete.
3. Click the Delete button to remove that client.
Note: Deleting a client does not erase that client' s index, so that client backup data can remain retrievable until the retention period expires.
Scheduling Large Server File Systems
Assuming a backup rate of 150 kilobytes per second, it takes almost 19 hours to fully back up a server that held 10 gigabytes of data. If you want an EBR service to run automatic, unattended backups, you may not have this much time. If you have servers with very large disk volumes to back up, you can split them into smaller sets and back them up on different schedules.
By segregating the server's disk volumes into different backup groups, you will still be able to protect the client's files without trying to do a full backup of the entire server, which will take a long time and require many tape changes.
Use the Save set display in the Editing Client dialog box to list the volumes or directories EBR backs up for the server. By default, all the disk volumes are backed up, as long as the keyword All is left in the list.
You segregate the disk volumes by adding the server's name twice to the EBR service's list of clients to back up. You then choose different schedules and save sets for each.
Important: If you are using save set scheduling and you create any services that do not have their data directory on /disk1, you must be sure that /disk1 is backed up before the disk on which the service's data directory reside. Otherwise, the backup data for the service will not be accessible through the Console Recover Browser.
The example that follows illustrates how to separately schedule backups for two server disks. The disk volumes will be separated so that:
/disk1
will be backed up according to a "Monday Full" schedule, and
/disk2
will be backed up according to a "Wednesday Full" schedule. (For this example, assume the EBR service already has custom schedules called "Monday Full" and "Wednesday Full" that were previously created.)
Important: Make sure you use the correct path when you specify volumes and directories. An EBR service cannot back up files when given incorrect paths. Use a path of either All or /diskn, where n is the number of the disk volume you want to back up.
To configure the client, complete the following steps:
1. Choose Setup from the Client menu, and click Create. The Create Client dialog box appears.
2. Type the name of the server into the Name field.
3. Configure the client by choosing backup group and directive. Make sure you choose the "Monday Full" schedule for it.
4. In the Save set field, delete "All" and enter "/disk1."
5. Click OK to save your changes.
Follow the same procedure for the remaining disk volume you want to back up on the "Wednesday Full" schedule:
1. Choose Setup from the Client menu, and click Create. The Create Client dialog box appears.
2. Type the Name of the server again, so as to schedule a backup for the second volume.
3. Configure the client by choosing backup group and directive. Make sure you choose the "Wednesday Full" schedule for it.
4. In the Save set field, delete "All" and enter "/disk2" instead.
5. Click OK to save your changes.
The Clients dialog box now lists the client twice, with different save sets to distinguish the two. The EBR service will back up the server' s disk volumes on two separate schedules.
Important: When you use the Save set scheduling feature, make sure you do not leave any part of the file system off the Save set list. Any disk volumes, directories, or files left off the list will never be backed up. All VINES servers include at least two file systems: the root file system in "/" and the "/disk1" file system.
Volume pools allow you to sort your backup data to selected volumes. A volume pool is a collection of volumes containing specific data sorted during the backup process. All EBR volumes belong to a pool - either the Default pool, the Archive pool, or a pool you have chosen.
You can sort data into a volume pool by selecting the type of backup data to be included in the pool. EBR uses the choices you make with the Pools command to sort the backup data to specific backup volumes.
Leaving a choice for Groups, Clients, Save sets or Levels un-selected in the Edit Pool dialog box means that the EBR service will not sort the data by that choice. For example, if you do not choose a level from the Levels choices, the EBR service will not sort the data by levels for that pool. EBR will sort the data based on other selections for that pool. You can sort data by:
![]()
A backup group or client ![]()
A list of save sets (file systems) ![]()
Backup levels - fulls, levels 1 to 9, incremental, or manual ![]()
Archive data only (for clients using Archive)
You can further organize your backup data by:
![]()
Choosing a special label template ![]()
Sending all backup data to a particular backup device ![]()
Choosing to store entries for the files in the on-line index (or not, if you are sending the backup media to an off-site vault)
Volume pools establish a logical and systematic method for tracking and organizing EBR backup volumes. For example, you may want to create a separate pool for each department in your company, or a pool containing only files with confidential information. You may want to assign all your full backups to a particular pool so you can easily store them off-site. Volume pools allow you to back up specific files or data onto specific media volumes.
Using the Pools command, it is easy to create a volume pool. However, you cannot create a new pool if EBR is busy doing a backup. This precaution ensures that EBR sends the backup data to the correct pool. In addition, you cannot delete a volume pool if any backup volumes still belong to that pool.
Unless you issue the Pools command, EBR backs up all clients in a group according to a schedule, and mixes the clients' data onto backup volumes. EBR tracks the backed-up data and the volumes to which data was backed up. Data is combined on the backup volumes. If you issue the Pools command, EBR puts data from specific clients onto specific tapes.
Each pool has an accompanying label template that you create in the Label Templates dialog box. You should use the same name for the label template as for the volume pool, to keep track of the pools to which backups belong. If you do not provide a label template while creating a pool, EBR automatically creates a template for you with the same name as the pool. The advantage to creating your own label template in the Label Templates window is that you have control over the label name.
To open the Pools window, choose Pools from the Media menu. The Pools dialog box appears:
The Pools dialog box displays existing volume pools, with the currently selected pool highlighted.
To modify an existing volume pool, complete these steps:
1. Highlight its name in the list.
2. Click the Edit button.
To add a new pool, click the Create button.
To eliminate a pool, click the Delete button. You may not delete a pool that contains backup volumes.
3. Highlight the Default pool.
4. Click the Edit button. The Edit Pool dialog box appears:
The Name field displays the name of the currently selected pool.
If you want the EBR service to use a pool during the backup process, select Yes from the Enabled choices. If you want to exclude a pool during the backup process, select No. When a pool is disabled, save sets intended for data that would normally go into that pool go to the Default pool instead.
Select a Label template from the choices. We recommend that you use the same name for the label template as for the pool.
The Groups field shows all groups known to the EBR server. There may be an unlimited number of groups in a pool.
You can use the Clients field to add individual client(s) to a pool without selecting a group, or to pick a specific client belonging to a group. If you pick a specific client from a group you will exclude all the other clients from that group. If you have a group selected, the client you select must belong to that group.
You can use the Save sets field to:
![]()
Send a specific file system to a pool. ![]()
Send a specific file system, from a specific client, to a volume pool without selecting a group (see "Example 2: Secure Device" in this chapter for an example) ![]()
Send a specific file system, from a specific client of a group, to a volume pool (see "Example 1: Separate Departments" in this chapter for an example)
The Archive only field should be set to Yes for archived save sets (the Archive pool), and No for other volume pools.
The Devices choices display the devices recognized by the EBR service. This is where you choose which devices to use for the selected pool. If you do not choose a device, EBR uses any available device.
Store index entries allows you to choose whether or not you want to include the index entries from the backup in the on-line index. For example, if you create a pool for Offsite, you may not want file index entries stored on-line, because a user request to recover data from this volume requires you to retrieve the volume from offsite storage. However, media index entries are still made.
Click the OK button to apply selections you made for a particular pool.
EBR is shipped with five pre-configured pools and corresponding label templates. Each pre-configured volume pool comes with a set of choices. If you do not choose a pool for your backups, they will automatically be assigned to the Default pool and labeled with the Default label template.
You can use the Default and Archive pools without changing them. To use other pre-configured pools, you must first complete some selections and choose Yes from the Enabled choices. A pool must be enabled in order for EBR to sort data to that pool.
The five pre-configured pools are:
![]()
Archive - For archiving client data only. This pool cannot be modified or deleted. There are no selections for you to make in this pool. ![]()
Default - Used automatically if you do not choose a pool. If you decide not to use the pools feature, EBR automatically places all of your backup volumes in this pool. The Default pool cannot be modified or deleted. There are no selections for you to make in this pool. ![]()
Full - For full backups only. This pool separates full backups from incremental and differential level backups. ![]()
NonFull - For any backups other than full backups. This pool includes all incremental and differential level backups. ![]()
Offsite - For volumes being stored off-site. If on-site backup volumes are destroyed, you could still recover data with the volumes stored off-site. If you are using the Full pool, you must temporarily disable it while sending data to the Offsite pool, to ensure that all full backups go only to the Offsite pool.
The Full and NonFull pools are intended for sorting data by levels, and are typically used in conjunction with each other.
Important: Remember to enable the pools you wish to have in effect during the scheduled backups by selecting Yes from the Enabled choices.
You can create as many pools as you need to effectively organize and sort your backup volumes. This section gives an example of a new pool.
Note: EBR Administrator does not allow you to create two different pools with duplicate choices for Groups, Clients, Save sets, Levels, or Archive Only in the Pools window. If you try to create two pools with the same choices for all of these items, EBR Administrator displays an error message.
Suppose you want to create a pool named "Test" that selects all the full backups from a group called QA. The QA group backs up all the data from systems in the quality assurance lab. You want to sort the full backups so you can have a complete set of backup tapes to store in a secure vault. You plan to keep data for a year in case it is ever needed, but do not anticipate a need to recover it.
Before creating a new pool, you need to create a label template with the same name - for example, Test.
To create a label template, complete the following steps:
1. Choose Label Templates from the Media menu. The Label Templates dialog box appears:
2. Click the Create button to devise a label template called Test. The Create Label Template dialog box appears.
3. Fill in the blanks, as shown in the following graphic:
4. Click the OK button. The Test label template appears as a choice in the Label Templates dialog box.
After creating the label template, open the Pools dialog box and follow these steps to create a pool named Test:
1. Click the Create button and enter "Test" into the Name field.
2. Select Yes from the Enabled choices.
3. Choose the label template Test, which you just created.
4. Choose the group QA, because it contains the clients you want to include in the Test pool. You do not want to selectively include clients or save sets, so do not use either the Clients or Save sets fields.
5. Select full from the levels choices. Leave the Archive only choice as No.
6. Select No from the Store index entries display. Selecting No means that you do not be include the index entries in the on-line index, which saves space on your hard disk.
7. Click the OK button to apply the selections.
The Create Pool dialog box should now look like this:
Once you create a new pool you can easily modify it by choosing the pool from the Pools list and changing the current selections. When you have completed making the modifications, click the OK button. If you have changed your mind and do not want to implement the new selections, click the Cancel button to restore the Pools window to its original configuration.
This section provides some examples of volume pools.
Example 1: Separate Departments
Your company needs to track department resources for accounting requirements that include the related costs of each department's backups. To sort each department's data, create a separate pool for each one. Sorting the data into different pools for each department makes it easy for accounting to determine how many tapes are being used, which backup devices are being used, and how much time is being spent on administering each department` s backup needs.
To create a pool for each department, follow these steps:
1. Create a label template with the name of the department in the Label Templates window.
2. Use the Pools window to create a new pool with the same name as the label template.
3. Select the groups that belong to each department.
You have confidential company information that must be backed up to a secure device in an area that only authorized personnel can access. You do not want to include an entire group, but just a select set of files from a small number of clients in a pool named "Confidential."
To create the "Confidential" pool, follow these steps:
1. Create a label template named "Confidential."
2. Create a pool with the same name.
3. Enter the name of each client in the Clients field.
4. Enter the file systems for each of the clients in the Save sets field.
5. Select the device they will be backed up to from the Devices choices.
Example 3: Targeted Backups
Your company creates large multi-gigabyte graphic files that require many backup volumes to complete a full backup. You want to back up these files only to the 8mm tape drive on your network, so you do not need to load new backup volumes as they fill up with data.
To back up large multi-gigabyte graphic files, follow these steps:
1. Create a label template named "Graphics Full."
2. Create a new pool by the same name.
3. Choose the group or groups that need to be included.
4. Select full from the Levels choices.
5. Choose the 8mm device from the Devices choices where the graphic files will be backed up automatically.