[Contents] [Prev. Chapter] [Next Section] [Next Chapter] [Index] [Help]

10    Managing ASE Services

This chapter describes how to use the asemgr utility to manage the services in an available server environment (ASE). Specifically, you can perform the following tasks with the asemgr utility:

[Contents] [Prev. Chapter] [Next Section] [Next Chapter] [Index] [Help]

10.1    Displaying Service Status

To display the status of available server environment (ASE) services, choose the "Display the status of a service" item from either the Managing ASE Services menu or the Obtaining ASE Status menu. Service status information includes the following:

Example 10-1 shows how to display the status of the NFS service named ase4.

Example 10-1:  Displaying Service Status

		Managing ASE Services
 
    c)  Service Configuration    
-->
    r)  Relocate a service
   on)  Set a service on line
  off)  Set a service off line
  res)  Restart a service
    s)  Display the status of a service
    a)  Advanced Utilities       
-->
 
    x)  Exit to the Main Menu            ?)  Help
 
Enter your choice [x]: s
 
 
        Service Status
 
Select the service whose status you want to display:
 
    1)  ase4 on toto
    2)  aseba2 on daffy
    3)  disk1  on gideon
 
    x)  Exit                             ?)  Help
 
Enter your choice [x]: 1
 
        Status for NFS service `ase4`
 
 Status:      Relocate:    Placement Policy:    Favored Member(s):
 on toto      yes          Balance services     None
 
        Storage Configuration for NFS service `ase4`
 
NFS Exports list
 /nfstest tregtest
 
Mount Table (device, mount point, type, options)
 treg#fset1 /var/ase/mnt/ase4/nfstest advfs rw
 
Advfs Configuration
 Domain:          Volume(s):
 treg             /dev/vol/dg3/vol01
LSM Configuration
 Disk Group:      Device(s):
 dg3              rz19g rz27g

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.2    Displaying Service Information from the ASE Database

You can obtain the full contents of the available server environment (ASE) database (/var/ase/asecdb) from disk in ASCII format by selecting the "Display the configuration of the ASE database" item from the Managing the ASE menu of the asemgr utility. Similarly, you can obtain the contents of the database for any given service by selecting the "Display the configuration of a service" item from the "Service Configuration" item of the Managing ASE Services menu.

Inspecting the contents of the ASE database can be helpful when you are staging a new ASE service, verifying the specific details of a service, or debugging unexpected service behavior. Because the format of the database output is in ASCII format, and tokens delimit the start and end of each action script, you can write scripts (such as a service monitor) that process it.

The asemgr command provides command-line flags that also allow you to dump ASE database contents. To display the information for a single ASE service, use the following command-line syntax:

asemgr -d -c [service ]

To display the information for the current ASE database, use the following syntax:

asemgr -d -C

By default, the asemgr utility dumps the contents of the ASE database from the /var/ase/config/asecdb file. To display the information in a specific ASE database (for example, a saved version), use the following syntax:

asemgr -d -C /var/ase/config/asecdb.backup

The ASE database on disk (rather than the one in memory) is read by these commands. The output from these commands can be read by a shell or awk script and parsed into actions, depending on the data.

For example, to dump the description of the drd1 service, enter the following command: 

# asemgr -d -c drd1
 
!! ---------------------------------------------------------------------------
!! ASE service configuration for drd1
!! ---------------------------------------------------------------------------
 
@startService drd1
Service name: drd1
Service type: DRD
Relocate on boot of favored member: yes
Placement policy: balanced
 
DRD Device Special File:    /dev/rdrd/drd1
Underlying Storage:         /dev/rvol/dg1/vol01
LSM disk group: dg1
  dg1 disks: rz10 rz9
@endService drd1

The following is an example of the type of information displayed when you enter an asemgr -d -C command or select the "Display the configuration of the ASE database" item from the Managing the ASE menu. Callout numbers refer to the descriptions listed after the example. 

                Managing the ASE
 
    a)  Add a member
    d)  Delete a member
    n)  Modify the network configuration
    m)  Display the status of the members
    C)  Display the configuration of the ASE database
    l)  Set the logging level
    e)  Edit the error alert script
    t)  Test the error alert script
     )  Enable ASE V1.5 functionality
 
    q)  Quit (back to the Main Menu)    
    x)  Exit                             ?)  Help
 
Enter your choice [q]: C
 
!! ---------------------------------------------------------------------------
!!  ASE database /var/ase/config/asecdb
!! --------------------------------------------------------------------------- [1]
 
ASE functionality version: V1.5 [2]
ASE logging level: Notice [3]
Number of ASE members: 2 [4]
Primary network default retry settings: (see /etc/hsm.conf for overrides) [5]
  Primary network max retries: 4 [6]
  Primary network time between retries: 10 [7]
  Primary network time between successes: 30 [8]
  Primary network time between failures: 50 [9]
Backup network default retry settings: (see /etc/hsm.conf for overrides) [10]
  Backup network max retries: 4
  Backup network time between retries: 10
  Backup network time between successes: 300
  Backup network time between failures: 300
AM default retry settings: (see /etc/hsm.conf for overrides) [11]
  AM max retries: 2
  AM time between retries: NA
  AM time between successes: 20
  AM time between failures: 300
Network interface default retry settings: (see /etc/hsm.conf for overrides) [12]
  Network interface max retries: 4
  Network interface time between retries: 10
  Network interface time between successes: 30
  Network interface time between failures: 50
 
!! ---------------------------------------------------------------------------
!! ASE database information
!! ---------------------------------------------------------------------------
 
ASE database format: V1.5 [13]
ASE database created by product: TruCluster Production Server [14]
ASE database created by member: lildogrm [15]
ASE database timestamp version: 870356729 [16]
ASE database last updated on: Thu Jul 31 09:45:29 1997 [17]
ASE database last updated by: lildogrm [18]
ASE database revision number: 16 [19]
 
!! ---------------------------------------------------------------------------
!! ASE network interface configuration
!! ---------------------------------------------------------------------------
 
Member name: lildogrm [20]
  Member IP address: 10.0.0.2 [21]
  Daemon communication use: Primary [22]
  Member aliveness ping use: Primary [23]
  Monitor interface setting: ignore [24]
Member name: lilcatrm [20]
  Member IP address: 10.0.0.1 [21]
  Daemon communication use: Primary [22]
  Member aliveness ping use: Primary [23]
  Monitor interface setting: ignore [24]
 
!! ---------------------------------------------------------------------------
!! ASE alert script information [25]
!! ---------------------------------------------------------------------------
 
System alert script name: /var/ase/sbin/ase_run_sh [26]
System alert script timeout: 20 [27]
 
@startText alert_script_0 [28]
# *****************************************************************
# *                                                               *
# *    Copyright (c) Digital Equipment Corporation, 1991, 1997    *
# *                                                               *
# *   All Rights Reserved.  Unpublished rights  reserved  under   *
# *   the copyright laws of the United States.                    *
# *                                                               *
# *   The software contained on this media  is  proprietary  to   *
# *   and  embodies  the  confidential  technology  of  Digital   *
# *   Equipment Corporation.  Possession, use,  duplication  or   *
# *   dissemination of the software and media is authorized only  *
# *   pursuant to a valid written license from Digital Equipment  *
# *   Corporation.                                                *
# *                                                               *
# *   RESTRICTED RIGHTS LEGEND   Use, duplication, or disclosure  *
# *   by the U.S. Government is subject to restrictions  as  set  *
# *   forth in Subparagraph (c)(1)(ii)  of  DFARS  252.227-7013,  *
# *   or  in  FAR 52.227-19, as applicable.                       *
# *                                                               *
# *****************************************************************
# @(#)$RCSfile: ase_logcrit_sh.sh,v $ $Revision: 1.2.5.3 $ (DEC) 
# $Date: 1996/06/26 21:19:55 $
 
ADMIN="root"
 
PATH=/sbin:/usr/sbin:/usr/bin
export PATH
 
TIME=`date +"%D %T"`
 
ERR_FILE=/var/ase/tmp/alertMsg
HSM_STATUS=`awk -F: '{print $2}' ${ERR_FILE} | sed 's/ //g'`
 
case    "${HSM_STATUS}" in
                HSM_PATH_STATUS)
                        awk -f /var/ase/lib/path_status_awk ${ERR_FILE}
                        ;;
                HSM_NI_STATUS)
                        awk -f /var/ase/lib/ni_status_awk ${ERR_FILE}
                        ;;
esac
 
if [ -n "${ADMIN}" ]; then
      if [ ! -f "${ERR_FILE}" ]; then
        echo "Critical ASE error or status change detected on `date`" > ${ERR_FILE}
      fi
 
        mailx -s "***Critical ASE error or status change - ${TIME}" ${ADMIN} < ${ERR_FILE}
fi
 
rm -f ${ERR_FILE}
 
 
@endText alert_script_0 [29]
 
 
!! ---------------------------------------------------------------------------
!! ASE service configuration for drd_service_1
!! ---------------------------------------------------------------------------
 
@startService drd_service_1 [30]
Service name: drd_service_1 [31]
Service type: DRD [32]
Relocate on boot of favored member: yes [33]
Placement policy: favored [34]
Favored member(s): milesd[35]
DRD Device Special File:    /dev/rdrd/drd1 [36]
Underlying Storage:         /dev/rrz19c [37]
@endService drd_service_1 [38]
 
 
!! ---------------------------------------------------------------------------
!! ASE service configuration for disk_service_1
!! ---------------------------------------------------------------------------
 
@startService disk_service_1 [30]
Service name: disk_service_1 [31]
Service type: DISK [32]
Relocate on boot of favored member: yes [33]
Placement policy: favored [34]
Favored member(s):  milesd[35]
Device: /dev/rz20c [39]
  /dev/rz20c mount point: NONE/dev/rdrd/drd1
@endService disk_service_1 [38]
 
 
!! ---------------------------------------------------------------------------
!! ASE service configuration for aseqa236
!! ---------------------------------------------------------------------------
 
@startService aseqa236 [30]
Service name: aseqa236 [31]
Service type: NFS [32]
Relocate on boot of favored member: yes [33]
Placement policy: favored [34]
Favored member(s): [35]
NFS locking file: /var/ase/mnt/aseqa236/ase/aseqa236/.ase/nfs_lock [40]
IP address: 16.141.112.236 [41]
Device: dom1#set1 [39]
  dom1#set1 mount point: /var/ase/mnt/aseqa236/ase/aseqa236
  dom1#set1 filesystem type: advfs
  dom1#set1 mount options: rw
  dom1#set1 mount point mode: 777
AdvFS domain: dom1
  dom1 volumes: /dev/vol/dg1/vol01
LSM disk group: dg1
  dg1 disks: rz10 rz9
 
!! ---------------------------------------------------------------------------
!! NFS exports list for aseqa236
!! ---------------------------------------------------------------------------
 
@startText aseqa236_NFS_exports [42]
 /ase/aseqa236 [43]
@endText aseqa236_NFS_exports [44]
@endService aseqa236

  1. The exclamation points are comment characters. [Return to example]

  2. Identifies the version of ASE installed. [Return to example]

  3. Identifies the condition for which ASE logging is enabled. Values include:

    [Return to example]

  4. Identifies the number of members in the ASE [Return to example]

  5. Identifies network retry settings for the primary network, as defined in the /etc/hsm.conf file. See Section 1.2.3 for more information on the HSM daemon. [Return to example]

  6. The number of additional attempts to get a response from a member system before a member is determined to be down. [Return to example]

  7. The amount of time in seconds between attempts to get a response from a member system. [Return to example]

  8. The time in seconds between a successful network or SCSI interconnect ping to a member and the subsequent ping. [Return to example]

  9. The time in seconds between a failed network or SCSI interconnect ping to a member and the next attempt to ping the member. [Return to example]

  10. Identifies network retry settings for the backup network, as defined in the /etc/hsm.conf file. [Return to example]

  11. Identifies network retry settings for the Availability Manager (AM) driver, as defined in the /etc/hsm.conf file. See Section 1.2.4 for more information on the AM driver. [Return to example]

  12. Identifies network retry settings for the backup network, as defined in the /etc/hsm.conf file. [Return to example]

  13. Identifies the format version of the ASE database. [Return to example]

  14. Identifies the product that was used to create the ASE database. [Return to example]

  15. Identifies the ASE member that was used to create the database. [Return to example]

  16. The timestamp for the version of the ASE database. [Return to example]

  17. Identifies when the last update to the database was made. [Return to example]

  18. Identifies the ASE member that was used last to update the database. [Return to example]

  19. Identifies how many times the ASE database has been updated. [Return to example]

  20. Identifies the ASE member name. [Return to example]

  21. Identifies the ASE member's Internet Protocol (IP) address. [Return to example]

  22. Identifies the network path used by the ASE daemons. Values for this field include:

    [Return to example]

  23. Specifies whether ASE will use this network to send pings to members to determine whether they are alive. Values for this field include:

    [Return to example]

  24. Specifies whether ASE will monitor the interface. Values for this field include:

    [Return to example]

  25. The default error alert script which is called via alert. [Return to example]

  26. Identifies the script the ASE agent uses to invoke the alert script. [Return to example]

  27. Specifies the alert script's timeout value. [Return to example]

  28. Defines the start of a script. [Return to example]

  29. Defines the end of a script. [Return to example]

  30. Start token for service information. [Return to example]

  31. Name of the ASE service being defined. [Return to example]

  32. Type of service. Values include:

    [Return to example]

  33. Specifies whether the service should relocate to a more favorable member when such a member becomes available (for instance, it boots). [Return to example]

  34. Identifies the ASE placement policy. Values include:

    [Return to example]

  35. Identifies the members on which the service can run (that is, those that have been specified as favored or restricted). [Return to example]

  36. Identifies the DRD device special file corresponding to a DRD service. [Return to example]

  37. Identifies the raw device(s) underlying a DRD service. [Return to example]

  38. End token for service information. [Return to example]

  39. Details on the underlying storage, disk, Advanced File System (AdvFS), and Logical Storage Manager (LSM) for Network File System (NFS) and disk services. [Return to example]

  40. NFS locking file for NFS services. [Return to example]

  41. IP address of disk, tape, or NFS service. [Return to example]

  42. Start token for NFS exports list for NFS services. [Return to example]

  43. List of exported NFS file systems. [Return to example]

  44. End token for NFS exports list for NFS services. [Return to example]

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.3    Placing a Service Off Line

You can use the asemgr utility to temporarily stop a service by placing it off line. While a service is off line, it is unavailable to clients.

When you place a Network File System (NFS) service off line, the file systems are automatically unmounted. When you place the service on line to start it, the file systems will be mounted.

When you place off line a service that uses the Logical Storage Manager (LSM), the disk groups are deported. Deporting the disk groups only makes them inaccessible; the disk groups, the volumes, and the data in the volumes are not deleted. When you set the service on line to start it, the disk groups will be imported.

To temporarily stop a service, choose the "Set a service off line" item from the Managing ASE Services menu and choose the service you want to place off line.

If you try to set a service off line and the service cannot be stopped, the asemgr utility displays the following message: 

ASE was unable to stop service `disk1`.  Check the syslog's
daemon log to determine why the stop action failed.
 
Two common reasons for the stop action to fail are:
 
(1) One of the service's filesets is in use
    (`umount` fails with Device Busy error)
 
(2) The user-defined stop script returns an error
 
You can fix the problem now and let ASE try to stop the
service, or you can ignore this failure and let ASE take the
service off line.
 
Enter 'r' for ASE to RETRY the stop action or 'o' take
the service OFFLINE [r]:

If you choose to retry the stop action by choosing r and it is successful, the service is placed off line. If you retry the stop action and it is unsuccessful, the asemgr utility displays the previous message again.

If you choose to continue the offline operation by choosing o, the asemgr utility sets the service off line. You must then manually stop all service processes, unmount the file systems or filesets, deport any imported LSM disk groups, and set the LSM disks off line.

If you added a user-defined stop script to the service, you must also manually perform the steps specified by this script. You can use the -d and -c options with the asemgr utility to dump the description of an ASE service, including the contents of any user-defined stop script (see Section 10.2).

To start a service that has been temporarily stopped, choose the "Set a service on line" item from the Managing ASE Services menu and then choose the service that you want to place on line.

The following list summarizes actions taken by the asemgr utility to stop a disk, Network File System (NFS), or shared tape service. If you must manually stop a service, perform these steps in sequence.

To stop a distributed raw disk (DRD) service, enter the following command:

/var/ase/sbin/drdmgr 0 stop service

The 0 directs errors to the /var/ase/tmp/ChildLog.0 file.

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.4    Restarting Unassigned Services

If the status of a service is unassigned, you can manually restart the service. To do this, choose "Restart a service" item from the Managing ASE Services menu and then choose the service you want to restart.

When you restart an unassigned service, the TruCluster software performs the following steps to completely stop the service:

  1. Executes any user-defined stop script for the service.

  2. Removes the Internet Protocol (IP) alias for the service.

  3. Unmounts file systems, tapes, raw disks, or Advanced File Server (AdvFS) filesets used by the service.

  4. Stops Logical Storage Manager (LSM) volumes and deports associated disk groups.

  5. Unreserves disks.

If this sequence completes successfully, the TruCluster software attempts to start the service. If any of the steps to stop a service fails, the TruCluster software does not attempt to start the service, giving you an opportunity to fix the problem before you manually restart the service. For example, LSM volumes may be left configured on the system because the step to deconfigure LSM did not run. The asemgr utility will tell you that the stop has failed. Look at the daemon.log file to see if the disk groups were deported. (You can also use LSM tools from the Cluster Monitor to get this information.) If cleanup is required, you can do this yourself or use the asemgr utility to place the service off line, which performs cleanup for you.

If the TruCluster software encounters a problem when trying to start a service after completing the stop sequence, the stop scripts are run again to try to clean up the problem. This has the same effect as placing the service off line, except that the status does not show as off line when you use the asemgr status menu.

When a service restart fails, look in the daemon.log file for error messages, which may help you determine why a service was unable to start. If possible, fix the problem that prohibited the service from being started and try again to restart it. You may need to place the service off line to get to a state so that the service can be successfully restarted.

Note

If an ASE service becomes unassigned to a member system, you may not be able to use the "Restart a service" item from the Managing ASE Services menu to bring the service back on line after the error has been corrected. To make the service operational, you may need to use the "Set a service off line" and "Set a service on line" items to disable and reenable the service.

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.5    Manually Relocating a Service

You can use the asemgr utility to manually relocate a service. For example, you may want to relocate a service if the member system currently providing the service needs maintenance. When you relocate a service, the TruCluster software stops the service on the member currently running the service and starts the service on another member.

When you use the asemgr utility to relocate a service, you can override the service's Automatic Service Placement (ASP) policy, which may restrict the service to specific members. See Chapter 4 for more information on ASP policies.

To relocate a service, follow these steps:

  1. Choose the "Relocate a service" item from the Managing ASE Services menu.

  2. Choose the service you want to relocate.

  3. Choose the member system that you want to run the service.

When you know the name of the service you want to relocate, you can do it from the command line. For example, to relocate the service named disk1 to member gideon, enter the following command: 

 # asemgr -m disk1 gideon

Example 10-2 shows how to relocate a service using the menu interface.

Example 10-2:  Relocating a Service

Select the service you want to relocate:
 
Services:
 
    1)  aseba1 on daffy
    2)  aseba2 on gideon
    3)  disk1 on toto
 
    x)  Exit to Managing ASE Services        ?)  Help
 
 Enter your choice [x]: 2
 
Select member to run 'aseba2' service:
 
    1)  toto
     )  gideon
    3)  daffy
 
    x)  Exit without making changes      ?)  Help
 
 Enter your choice: 1
 
relocating service `aseba2` to member`toto`...
relocation successful...
 

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.6    Modifying a Service

The "Service Configuration" item of the Managing ASE Services menu of the asemgr utility provides two options that allow you to change service parameters in the ASE database.

You can make some changes to a service while it remains on line to its clients by selecting the "Modify a service without interrupting its availability" item, and following the instructions in Section 10.6.1. Other changes require minimal interruption of a service's availability to clients. To perform these operations, select the "Modify a service" item, and follow the instructions in Section 10.6.2.

Table 10-1 lists indicates which operations can be performed from the "Modify a service without interrupting its availability" menu item. You can perform all of the operations listed from the "Modify a service" item, but the asemgr utility will put the service off line while it accomplishes them.

Table 10-1:  Service Modification Impact on Service Availability

Operation "Modify a service without interrupting its availability" "Modify a service"

Change the Automatic Service Placement (ASP) policy--See Chapter 4 for information about the ASP policies you can assign to a service.

Yes

Yes

Update the ASE database to reflect the addition of raw devices (/dev/rz*, or /dev/*mt* for tape services) to, or the removal of raw devices from, a Logical Storage Manager (LSM) disk group participating in a disk-based service.

Yes

Yes

Update the ASE database to reflect the addition of raw devices to, or the removal of raw devices from, an Advanced File System (AdvFS) domain participating in a disk-based service.

Yes

Yes

Update the ASE database to reflect the addition of LSM volumes (/dev/vol/*/vol*) to, or the removal of LSM volumes from, an AdvFS domain participating in a disk-based service.

Yes

Yes

Modify the exports file for a Network File System (NFS) service.

Yes

Yes

Add raw devices to, or remove raw devices from, a distributed raw disk (DRD) service.

Yes

Yes

Add LSM volumes (/dev/vol/*/vol*) to, or remove LSM volumes from, a DRD service.

Yes

Yes

Add UNIX file systems or AdvFS filesets to a disk-based service.

No

Yes

Change the name of a file system, fileset, or volume participating in a service.

No

Yes

Change the mount point of a service, including its mount options, owner, and mode.

No

Yes
Change the disk access mode (either read/write or read-only) or quotas.

No

Yes

Change a service's name.

No

Yes

Replace, edit, modify, or delete a user-defined action script.

No

Yes

Change the timeout value for a user-defined action script.

No

Yes

DIGITAL recommends using either of these options whenever you change a service's configuration information in the ASE database. Both allow the service to remain on line in its ASE after the modifications are performed expeditiously under ASE control.

You can also modify a service while it is off line to the ASE (for example, if you have set it off line as described in Section 10.3) and outside of the utility's control. However, this method requires more manual intervention, is most disruptive to service availability, and is prone to error, especially in complex storage configurations.

Notes

Carefully read Section 10.6.3 before modifying AdvFS or LSM configuration information for a service that you have set off line. This section also describes restrictions to the use of AdvFS, LSM, and DIGITAL UNIX disk management commands in an ASE.

If you modify an NFS service in an ASE, your first attempt to access the service may cause a stale file handle message to be displayed

After you modify the exports file of an NFS service, clients may receive stale file handle messages and be unable to mount NFS file systems associated with the service. If this problem occurs, send the SIGHUP signal to the mountd process on the ASE member that is running the service.

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.6.1    Modifying a Service Without Interrupting Its Availability

For those changes to a service that do not require a disruption in service availability (as described in Table 10-1), follow these steps:

  1. Run the asemgr utility on the member system on which the service is currently running.

  2. Choose the "Modify a service without interrupting its availability" item from the Service Configuration menu and then choose the service that you want to modify: 

     Service Configuration
 
   a)  Add a new service
   m)  Modify a service
   o)  Modify a service without interrupting its availability
   d)  Delete a service
   s)  Display the status of a service
 
   x)  Exit to Managing ASE Services    ?)  Help
 
Enter your choice [x]: o
 
 Online Service Modification
 
Select the service you want to modify:
 
   1)  ase1 on fgreg1
   2)  greg on fgreg2
   3)  drd1 on fgreg2
 
   x)  Exit to Service Configuration
 
Enter your choice [x]: 2

    The asemgr utility displays a list of modifications that can be performed without disrupting the selected online service.

  3. For example, choose option a to modify the Automatic Service Placement (ASP) policy: 

    Select what you want to modify in service `greg`:
 
    a)  Automatic service placement (ASP) policy
    u)  Update with storage configuration changes (LSM and AdvFS only)
 
    x)  Exit without modifications
 
Enter your choice [x]: a

    At this point, proceed to specify a new ASP policy in the same manner as when adding a new service. For more information on specifying an ASP policy, see Chapter 4.

Example 10-3 shows how to update disk configuration information for an online AdvFS service. In this example, the administrator added a new volume to an AdvFS domain and runs the asemgr utility to modify the service (ase1) that uses this domain. The asemgr utility checks the AdvFS configuration information and updates the ASE database while service ase1 continues to run without interruption.

Example 10-3:  Modifying an AdvFS Service with No Disruption to Its Availability

		Service Configuration
 
    a)  Add a new service
    m)  Modify a service
    o)  Modify a service without interrupting its availability
    d)  Delete a service
    s)  Display the status of a service
 
    x)  Exit to Managing ASE Services    ?)  Help
 
Enter your choice [x]: o
 
	Online Service Modification
 
Select the service you want to modify:
 
    1)  ase1 on fgreg1
    2)  greg on fgreg2
    3)  drd1 on fgreg2
 
    x)  Exit to Service Configuration    
 
Enter your choice [x]: 1
 
Select what you want to modify in service `ase1`:
 
    a)  Automatic service placement (ASP) policy
    u)  Update with storage configuration changes (LSM and AdvFS only)
 
    x)  Exit without modifications       
 
Enter your choice [x]: u
 
Checking AdvFS domain val ...
 
AdvFS domain `val` shows a change in the volumes configured:
 
Old volume list:
	/dev/rz20b
 
New volume list:
	/dev/rz16b
	/dev/rz20b
 
Is this correct (y/n) [y]: y
Enter 'y' to modify service 'ase1' (y/n): y
Service successfully updated.

If the service cannot be started with the new storage configuration information, the service is placed off line. You are then given the opportunity to correct the problem or to discard the configuration modifications.

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.6.2    Modifying a Service with Minimal Disruption to Its Availability

For those changes to a service that require a slight disruption in service availability (as described in Table 10-1), select the "Modify a service" item from the Service Configuration menu of the asemgr utility.

Figure 10-1 shows a map of the asemgr utility's Modifying a Service menu.

Figure 10-1:  Map for Modifying a Service Menu

When you select the "Modify a service" item from the Service Configuration menu of the asemgr utility, the TruCluster software automatically performs the following tasks:

  1. Stops the service on the member system.

  2. Deletes the service from all the members.

  3. Adds the service to the ASE database.

  4. Starts the modified service on a member.

  5. Propagates the database changes to all the members.

If you try to modify a service and the service cannot be stopped, the asemgr utility displays the following message: 

ASE was unable to stop service `disk`.  Check the syslog's
daemon log to determine why the stop action failed.
 
Two common reasons for the stop action to fail are:
 
(1) one of the service's filesets is in use
    (`umount` fails with Device Busy error), or
 
(2) the user-defined stop script returns an error.
 
You can fix the problem now and let ASE try again to stop the
service or you can let ASE take the service OFFLINE so you can
stop the service manually.  If you choose to stop the service
manually and cannot stop it completely, you will need to reboot
the system to avoid a system panic or data corruption.
 
If the stop action failed because one of the service's filesets
is in use, you should try to kill the processes that are using
the fileset and let ASE try again to stop the service.  If ASE
is unable to stop the service, you should let ASE take the
service offline.  If the stop action failed because the
user-defined stop script returned an error, you should let ASE
continue with the modify operation.
 
 
Enter 'r' for ASE to RETRY the stop action or 'm' for you to
MANUALLY stop the service [r]:

If you retry the stop action and it is successful, the service is modified. If you retry the stop action and it is unsuccessful, the asemgr utility displays the previous message again.

If you continue the stop operation without fixing the problem, the asemgr utility displays the following message: 

You must manually stop all service processes, unmount any mounted
filesets, deport any imported LSM disk groups, and set the LSM
disks offline.
 
Failure to stop the service could cause the system to panic when
the service is restarted.  If you cannot stop the service, you
should reboot the member running the service.
 
 Press 'Return' to continue: [Return]
 
You can now exit the modify operation or continue with
the modify.  If you continue with the modify, the old service
will be deleted and then the new service will be added.  Once the
old service is deleted, if the ASE cannot add the new service or
restore the old service, the service will remain deleted.
 
Enter 'x' to exit the modify operation or 'c' to continue [x]:

The asemgr utility allows you to either continue to modify the service or to exit the operation. If you choose to exit the operation, the service is placed off line, and the asemgr utility returns to the Service Configuration menu. If you continue with the modification, the original service is deleted and the new service is added and then started, if possible.

If a modified service cannot be added to the ASE, the asemgr utility displays the following message: 

Add failed - Unable to add service.
Check syslog's daemon log to determine the error.
 
Enter 'o' to restore the old service configuration, 'n' to retry
the new service configuration, or 'd' to delete the service [n]:

For example, the mount point specified in an NFS service may not have been created on the system. You can create the mount point, then type n to retry.

If you retry the service modification and it succeeds, the modification proceeds. It the retry fails, the asemgr utility displays the previous message.

If you successfully restore the original service, it remains configured but is placed off line. If the original service cannot be restored, the asemgr utility displays the previous message.

If you cannot successfully restore the original service or modify the service, you can delete the service. See Section 10.7 for information about deleting services.

Example 10-4 shows how to modify an NFS service by adding a new area to be exported, and by editing the ASE exports file to restrict the service to two clients.

Example 10-4:  Modifying an NFS Service

 # asemgr

.
.
.
Modifying a Service
 
Select the service you want to modify:
 
 
    1)  ase3
    2)  aseba2
 
    x)  Exit                             ?)  Help
 
 
 Enter your choice [x]: 1
 
Select what you want to modify in service `ase3`:
 
    g)  General service information
    r)  Automatic service placement (ASP) policy
 
    x)  Exit without modifications           ?)  Help
 
 Enter your choice [g]: g
 
                 NFS Service Modification
 
The following menu lists the disk storage configuration for
the service "ase3."
 
You can modify the following storage configurations, add more storage,
or perform miscellaneous modifications (for example, modify the
exports file).
 
Select what to modify in the NFS service `ase3`:
 
    1)  /dev/vol/dg3/vol03      (UFS)
    2)  dom1#fset1     (ADVFS)
    a)  Add a UFS file system, AdvFS fileset, or LSM volume
    m)  Miscellaneous modifications for `ase3`
    q)  Quit without making any changes
    x)  Exit (done with modifications)
 
 Enter your choice [x]: a
 
                   Specifying Disk Information
 
Enter one or more UFS device special files, AdvFS filesets, or LSM
volumes to define the disk storage for this service.
 
    For example:        Device special file:      /dev/rz3c
                        AdvFS fileset:            domain1#set1
                        LSM volume:               /dev/vol/dg1/vol01
 
To end the list, press the Return key at the prompt.
 
Enter a device special file, an AdvFS fileset, or an LSM volume as
 storage for this service (press 'Return' to end): /dev/rz28c
 
Enter the directory pathname(s) to be NFS exported from the storage
area "/dev/rz28c".  Press 'Return' when done.
 
 Enter a directory pathname: /usr/staff

  Enter a host name, NIS netgroup, or IP address for the NFS
   exports list (press 'Return' for all hosts): staff_group
 
Enter a directory pathname:[Return]
 
              UFS File System Read-Write Access
Mount `/dev/rz28c` file system with read-write or read-only access?
 
    1)  Read-write
    2)  Read-only
 
 Enter your choice [1]: 2
 
              UFS Mount Options Modification
 
Enter a comma-separated list of any mount options you want to use for
`/dev/rz28c` (in addition to the UFS-specific defaults listed
in the mount.8 reference page).  If none are given, only the default
mount options are used.
 
 Enter options (Return for none):[Return]
 
Enter a device special file, an AdvFS fileset, or an LSM volume as
 storage for this service (press 'Return' to end): [Return]
 
Select what to modify in the NFS service `ase3`:
 
    1)  /dev/vol/dg3/vol03      (UFS)
    2)  dom1#fset1     (ADVFS)
    a)  Add a UFS file system, AdvFS fileset, or LSM volume
    m)  Miscellaneous modifications for `ase3`
    q)  Quit without making any changes
    x)  Exit (done with modifications)
 
 Enter your choice [x]: m
 
Miscellaneous modifications for service `ase3`:
 
    e)  Export file input for `ase3`
    n)  Service name [ase3]
    s)  NFS file system area [dom1#fset1]
    u)  User-defined action scripts
    x)  Exit to previous menu
 
 
 Enter your choice [x]: x
 
Select what to modify in the NFS service `ase3`:
 
    1)  /dev/vol/dg3/vol03      (UFS)
    2)  dom1#fset1     (ADVFS)     
    a)  Add a UFS file system, AdvFS fileset, or LSM volume
    m)  Miscellaneous modifications for `ase3`
    q)  Quit without making any changes
    x)  Exit (done with modifications)
 
 Enter your choice [x]: x
 
 Enter 'y' to modify Service 'ase3' (y/n): y
 
Stopping service...
Deleting service...
Adding service...
Starting service...
Saving the updated database...
Service successfully updated...

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.6.3    Managing Storage Participating in ASE Services

Managing storage in an ASE requires a detailed understanding of the TruCluster software and storage subsystem operation.

This section provides information about managing AdvFS and LSM subsystems, replacing a failed disk, and using DIGITAL UNIX storage management commands.

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.6.3.1    Introduction to ASE Storage Configuration Management

Before you set up a service, you must have configured the disks that you will use in the service. For example, you must have set up your file systems, AdvFS filesets, and LSM volumes. When you run the asemgr utility to add a service, you specify information about the disk configuration at the prompts. This information is included in the ASE database on each member system.

However, after a disk is used in an ASE service, you must use special procedures to manage the disk. This is because the ASE must always maintain control of the disk while a service is running. For example, you do not manually unmount a file system that is being used by an online ASE service.

As discussed in Section 10.6, the TruCluster software allows you to modify an ASE service and change its storage configuration. The "Service Configuration" item of the Managing ASE Services menu of the asemgr utility provides two options that allow you to change service parameters while the service remains on line within its ASE and under ASE control. You can modify any of the information that was specified when the service was added to the ASE. Although most common changes (such as adding a disk to an AdvFS domain or an LSM volume) can be made to a service without disrupting its availability to clients, others (such as changing disk attributes) can be made only with a slight disruption to its availability. Table 10-1 indicates which modifications require a disruption of service availability.

DIGITAL recommends using either of these options whenever you change a service's configuration information in the ASE database. Both allow the service to remain on line in its ASE while the modifications are performed expeditiously under ASE control. You can also modify a service while it is off line to the ASE and outside of the utility's control. However, this method requires more manual intervention, is most disruptive to service availability, and is prone to error, especially in complex storage configurations. Carefully read Section 10.6.3 before modifying AdvFS or LSM configuration information for a service that you have set off line. This section also describes restrictions on the use of AdvFS, LSM, and DIGITAL UNIX disk management commands in an ASE.

If you make modifications that affect a service's AdvFS or LSM storage configuration and you want to modify the service while it is off line to the ASE, you must set the service off line as described in Section 10.3 and follow the special procedures described in the following sections. These procedures ensure that the ASE recognizes and incorporates the configuration changes, and that the ASE database is consistent with any AdvFS and LSM subsystem databases.

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.6.3.2    Understanding Storage Configuration Databases

The AdvFS and the LSM subsystems utilize databases that contain records of the subsystems' disk configuration. AdvFS and LSM subsystems consist of a disk organization built on top of physical disks. For example:

Before you add an ASE service, you must set up the storage configuration for the service. When you create the service, the asemgr utility prompts you for information about the storage configuration, including physical disks, UNIX file systems, AdvFS filesets and domains, LSM volumes, and mount points. This information is included in the ASE database file on each member system and is used to configure, start, and stop the ASE service.

For proper ASE operation, the information in the ASE database must be consistent with any AdvFS or LSM configuration database. If you modify a service's AdvFS or LSM configuration and those modifications cause the subsystem database to change, you must be sure to incorporate those changes into the ASE database.

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.6.3.3    Modifying an Offline AdvFS or LSM Service

If you modify a service that has been placed off line, the service will be unavailable during the entire service modification procedure.

To modify an AdvFS or LSM storage configuration used in a service, you must follow special procedures:

Regardless of whether you are modifying an online or an offline service, some AdvFS and LSM storage modifications require you to update the ASE database with the storage configuration changes. You also must adhere to the AdvFS and LSM command restrictions listed in Section 10.6.3.4 and Section 10.6.3.5.

To modify the storage configuration for a service that has been placed off line, follow these steps:

  1. Optionally, run the asemgr utility to display the status of the service you want to modify and the storage configuration for the service.

  2. Run the asemgr utility and set the service off line.

  3. If you are modifying an LSM storage configuration or an AdvFS configuration on top of an LSM volume, perform the following tasks on the member system on which you will modify the LSM configuration:

    1. Place on line the disks used in the disk group you want to modify, using the following command syntax:

      voldisk online disk . . .

    2. Import the disk group, using the following command syntax:

      voldg import disk_group

    3. Restart the volumes, using the following command syntax:

      volrecover -sb disk_group

    If you are modifying an AdvFS storage configuration, perform the following tasks:

    1. On the member system on which you will modify the AdvFS configuration, re-create the domain directory, using the following command syntax:

      mkdir -p /etc/fdmns/ domain_name

    2. Change your directory to the domain directory, using the following command syntax:

      cd /etc/fdmns/ domain_name

    3. Re-create the device links, using the following command syntax for all the volumes in the domain:

      ln -s device

      The device variable can be a UNIX file system or an LSM volume (for example, /dev/rz20c or /dev/vol/dg3/vol04).

  4. Modify the AdvFS or LSM storage configuration. See Section 10.6.3.4, Section 10.6.3.5, and Section 10.6.3.6 for restrictions on using AdvFS, LSM, and DIGITAL UNIX disk management commands in an ASE.

  5. If you modified the LSM storage configuration, deport the disk group and take its disks off line. Use the following command syntax:

    voldg deport disk_group

    voldisk offline disk . . .

    If you modified the AdvFS storage configuration, remove the domain directory that you created, using the following command syntax:

    rm -rf /etc/fdmns/domain_name

  6. If necessary, update the ASE database by running the asemgr utility. Select the "Modify a service" or "Modify a service online" item, as appropriate, from the Service Configuration menu, specify the name of the service, and then select the "General service information" menu item.

    Choose the menu item that corresponds to the storage modification you made. For example: 

      Select what to modify in the NFS service `ase3`:
 
    1)  /dev/vol/dg3/vol03      (UFS)
    2)  dom1#fset1     (ADVFS)
    a)  Add a UFS file system, AdvFS fileset, or LSM volume
    m)  Miscellaneous modifications for `ase3`
    q)  Quit without making any changes
    x)  Exit (done with modifications)
 

    In the previous example:

    After you choose a menu item, another menu is displayed, depending on your previous choice. Select the menu item that corresponds to the storage modifications you made.

    If you chose the /dev/vol/dg3/vol03 item from the previous menu, the following menu might be displayed: 

    Modify/Delete file system `/dev/vol/dg3/vol03`:
 
    m)  Modify UFS file system information
    n)  Change file system device special file
           [/dev/vol/dg3/vol03]
    e)  Modify the exports list
     )  Modify AdvFS domain information
    d)  Delete `/dev/vol/dg3/vol03`
    x)  Exit - done with changes

    After you choose the appropriate menu item, information about the modified configuration is displayed, and you are prompted to confirm that the information is correct.

    If you chose item m in the previous example, the following information might be displayed: 

    LSM physical list does not match with what LSM thinks
 
        New list: rz10g rz11g
        Old list: rz10g
 
 Following is a list of devices and pubpath for disk group dg3
 
        DEVICE  PUBPATH
 
        rz10g   /dev/rz10g
        rz11g   /dev/rz11g
 
Is this correct (y/n) [y]:

    When you enter y, the TruCluster software updates the ASE database with the new storage configuration on all the member systems.

  7. Run the asemgr utility and place the service on line. If you are unable to restart the service because of an error, the service remains off line, giving you the opportunity to correct the problem.

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.6.3.4    Restrictions on Using AdvFS Commands in an ASE

The following list contains AdvFS commands that have restrictions when used in an ASE or that require you to update the ASE database:

The following list contains the AdvFS commands that have no restrictions when used in an ASE, and do not require you to update the ASE database:

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.6.3.5    Restrictions on Using LSM Commands in an ASE

The following list contains the LSM commands that have restrictions when used in an ASE, or that require you to update the ASE database:

The following list contains the LSM commands that have no restrictions when used in an ASE, and do not require you to update the ASE database:

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.6.3.6    Restrictions on DIGITAL UNIX Storage Management Commands

The following commands have restrictions when used in an ASE:

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.6.3.7    Replacing a Failed Disk

The following sections describe how to replace a failed disk. See your hardware configuration and software installation manuals for information about installing disks in storage units.

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.6.3.7.1    Replacing a Nonmirrored Disk

If a disk that is not part of an LSM mirrored volume fails, the service that is using the disk will stop. To replace a disk, follow these steps:

  1. Pull the failed disk from its slot and replace it with a disk that has the same unit number as the failed disk.

  2. If the disk is part of an LSM disk group, use the asemgr utility to rereserve the service's disks.

  3. Restore the data from a backup.

  4. Start the service.

If you are using LSM, you can replace a failed disk with a disk that has a different unit number. To do this, you must have a spare disk that is part of the same LSM disk group as the failed disk. Having the spare disk available ensures that the list of physical disks in the disk group does not change, and you do not have to update the ASE database. However, you must keep the failed disk in its disk slot until you are ready to replace it with a viable disk. This replacement disk can now be used as the spare disk in the disk group.

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.6.3.7.2    Replacing a Disk that is Part of an LSM Mirrored Volume

You can replace a failed disk that is part of an LSM mirrored volume without interrupting the availability of the service. To do this, follow these steps:

  1. Follow the procedure in the DIGITAL UNIX Logical Storage Manager manual for replacing a disk with the same unit number. However, while the failed disk is removed from the disk group, the service cannot be failed over.

  2. Run the asemgr utility and rereserve the service's disks.

  3. Use the volrecover command to recover the disk data.

You can replace a failed disk with a disk that has a different unit number without interrupting the availability of the service. To do this, you must have a spare disk that is part of the same disk group as the failed disk. Having the spare disk available ensures that the list of physical disks in the disk group does not change, and you do not have to update the ASE database. However, you must keep the failed disk in its disk slot until you are ready to replace it with a viable disk. This replacement disk can now be used as the spare disk in the disk group.

[Contents] [Prev. Chapter] [Prev. Section] [Next Section] [Next Chapter] [Index] [Help]

10.7    Deleting a Service

You use the asemgr utility to delete a service. When you delete a service, the TruCluster software stops the service on the member system, deletes the service from all the members, removes the service information from the database, and propagates the database changes to all the members.

If you delete or stop a service that uses Logical Storage Manager (LSM), the disk group is deported. Deporting the disk group only makes it inaccessible; the disk group, the volumes, and the data in the volumes are not deleted. In addition, if you delete or stop a service that uses Advanced File System (AdvFS), the domain is no longer configured. Because of this, when you delete a service that uses either LSM or AdvFS, the asemgr utility prompts you for a member on which to leave the LSM disk group imported or AdvFS domain configured, so you can use it for other purposes.

To delete a service, choose the "Delete a service" item from the Service Configuration menu and then choose the service you want to delete. If the service uses AdvFS or LSM, you are prompted for a member on which to keep the disks configured.

Example 10-5 shows how to delete a service.

Example 10-5:  Deleting a Service

 # asemgr
 

.
.
.
Service Configuration
 
    a)  Add a service
    m)  Modify a service
    d)  Delete a service
    s)  Display the status of a service
 
    x)  Back to Managing ASE Services menu   ?)  Help
 
 Enter your choice [x]: d
 
Select the service you want to delete:
 
    1)  aseba1 on daffy
    2)  aseba2 on gideon
    3)  disk1 on toto
 
    x)  Exit to Service Configuration     ?)  Help
 
 Enter your choice [x]: 2
 
This service uses either an AdvFS or an LSM storage configuration.
You must select a member on which to leave the storage configured:
 
    1)  toto
    2)  gideon
    3)  daffy
 
    x)  Exit to Service Configuration      ?)  Help
 
 Enter your choice [1]: 2
 
Member to leave the storage configuration on: gideon
 
 Is this correct (y/n) [y]: y
 
 Enter 'y' to delete Service 'aseba2' (y/n): y
 
Stopping service...
Deleting service...
Saving the updated database...
Service successfully removed...

If you try to delete a service and the service cannot be stopped, the asemgr utility displays the following message: 

ASE was unable to stop service `service`.  Check the syslog's
daemon log to determine why the stop action failed.
 
Two common reasons for the stop action to fail are:
 
(1) One of the service's filesets is in use
    (`umount` fails with Device Busy error)
 
(2) The user-defined stop script returns an error
 
You can fix the problem now and let ASE try to stop the service,
or you can ignore this failure and let ASE continue with the
delete operation.
 
Enter 'r' for ASE to RETRY the stop action or 'c' to CONTINUE
with the delete operation [r]:

If you retry the stop action and it is successful, the service is deleted. If you retry the stop action and it is unsuccessful, the asemgr utility displays the previous message again.

If you continue the delete operation, the asemgr utility displays the following message: 

You must manually stop all service processes, unmount any
mounted filesets, deport any imported LSM disk groups, and
set the LSM disks off line.
 
Failure to stop the service could cause the system to panic
when the service is restarted.  If you cannot stop the service,
you should reboot the member running the service.
 
Press 'Return' to continue:

When you press the Return key, the asemgr utility deletes the service. You must then manually stop all service processes, unmount any mounted file systems or filesets, deport any imported LSM disk groups, and set the LSM disks off line. You can then use the disks in another service or for some other purpose.

[Contents] [Prev. Chapter] [Prev. Section] [Next Chapter] [Index] [Help]

10.8    Rereserving an LSM Device

When a failed or previously unavailable part of a Logical Storage Manager (LSM) mirrored volume becomes available again, you can reincorporate the device into the service without interrupting the service. To do this, resynchronize the mirrored volume outside of the available server environment (ASE) on the member to which the disk groups are imported. Then, rereserve the devices by using the asemgr utility's Advanced Utilities menu.

[Contents] [Prev. Chapter] [Prev. Section] [Next Chapter] [Index] [Help]