Example output for restore performance monitoring
This topic presents an example of the output that is displayed from running the db2pd command with the -barstats option, for a running restore operation.
Where applicable, example sections are followed by a table that provides detailed descriptions of the fields that appear in the results.
Examples
First,
the agent and Engine Dispatchable Unit (EDU) identifiers for this restore operation are displayed.
The term EDU is synonymous with thread.
$ db2pd -db dbname1 -barstats 391
Printing out Restore Runtime Statistics at 2018-07-18-00.27.28.028659:
Restore Related EDUs:
---------------------------------------------------------------------
Restore agent ID: 391
BM 0 (EDU ID): db2bm.391.0 (488)
BM 1 (EDU ID): db2bm.391.1 (489)
MC 0 (EDU ID): db2med.391.0 (490)| Field | Description |
|---|---|
| Restore agent ID | Represents the EDU ID for the main restore agent, which coordinates the various backup EDUs. |
| BM N (EDU ID): db2bm.M.N | Represent the restore Buffer Manipulator EDU identifiers. These EDUs write data to table space container files. |
| MC N (EDU ID): db2bm.M.N | Represent the restore Media Controller EDU identifiers. These EDUs read data from one or more backup image files. |
Next, information about each table space to be restored is displayed. If an explicit list of
table spaces was specified in the restore command, then only those table spaces are restored and
listed here.
Table Spaces to be Restored (appTblSpace):
---------------------------------------------------------------------
numEntry: 1
Table Spaces:
tblSpaceName: SYSCATSPACE
tblSpaceID: 0
tblSpaceType: 2
tblSpaceDataType: 0
tblSpaceSize: 15568 (in 4 KB pages)
nContainers: 1
isInImage: T
doNotRestoreUserData: F
isPartOfRestore: T
isComplete: F
isLocked: F
tbspUseablePagesInImg: 16380
hwm: 0
hwmOverflow: F
tblSpaceName: USERSPACE1
tblSpaceID: 2
tblSpaceType: 2
tblSpaceDataType: 32
tblSpaceSize: 1824 (in 4 KB pages)
nContainers: 1
isInImage: T
doNotRestoreUserData: F
isPartOfRestore: T
isComplete: F
isLocked: F
tbspUseablePagesInImg: 4064
hwm: 0
hwmOverflow: F
| Field | Description |
|---|---|
| tblSpaceName | The name of the table space. (See tablespace_name). |
| tblSpaceID | The table space ID. (See tablespace_id). |
| tblSpaceType | The table space type. (See tablespace_type). |
| tblSpaceDataType | The table space content type. (See tablespace_content_type). |
| tblSpaceSize | The table space size (number of 4 KB pages). (See tablespace_usable_pages). |
| nContainers | The number of table space containers. (See tablespace_num_containers). |
| isInImage | A flag that indicates table space inclusion in the backup image. |
| doNotRestoreUserData | A flag that indicates whether data within the table space (aside from table space structural information) will be restored. |
| isPartOfRestore | A flag that indicates whether the table space is part of the restore set. |
| isComplete | A flag that indicates whether the restore operation is completed for the table space. |
| isLocked | A flag that indicates whether the table space is locked in preparation for the restore operation. |
| tbspUseablePagesInImg | The table space usable size (number of 4 KB pages). |
| hwm | The high water mark page of the table space. |
| hwmOverflow | A flag that indicates whether container size limit was exceeded while restoring the table space. |
Next, a list of table spaces that are excluded from the restore set is displayed. For example, if
a restore rebuild command runs as part of the operation, and an EXCLUDE table space list is
specified, then those table spaces are listed
here:
Table Spaces Excluded from Restore (exceptionTbspList):
---------------------------------------------------------------------
No table spaces.
Next, a list of all the table spaces that are included within the backup image is displayed:
Table Spaces Available in the Backup Image (imgTbspList):
---------------------------------------------------------------------
numEntry: 2
Table Spaces:
tblSpaceName: SYSCATSPACE
tblSpaceID: 0
tblSpaceType: 2
...etc...Next,
the restore performance statistical information is displayed, showing column data for the specific
restore operation. If a column is not applicable to a particular restore operation, it is not
displayed.
Performance statistics:
---------------------------------------------------------------------
Parallelism = 2
Number of buffers = 4
Buffer size = 6307840 (1540 4kB pages)
BM# Total I/O MsgQ WaitQ LogPhase Buffers kBytes LogBytes
--- -------- -------- -------- -------- -------- -------- -------- --------
000 0:00:13 0:00:12 0:00:00 0:00:01 0:00:00 14 86240 0
001 0:00:13 0:00:12 0:00:00 0:00:01 0:00:00 15 86268 0
--- -------- -------- -------- -------- -------- -------- -------- --------
TOT - - - - - 29 172508 0
MC# Total I/O MsgQ WaitQ Buffers kBytes
--- -------- -------- -------- -------- -------- --------
000 0:00:12 0:00:00 0:00:12 0:00:00 30 172512
--- -------- -------- -------- -------- -------- --------
TOT - - - - 30 172512| Field | Description |
|---|---|
| BM# | The Buffer Manipulator EDU identifier. |
| MC# | The Media Controller EDU identifier. |
| Total | The total elapsed time that this EDU has been in restore processing. |
| I/O | The amount of time that this EDU spent either reading or writing data. |
| MsgQ | The amount of time that the EDU spent waiting to get a buffer. For restore operations, the Media Controller EDUs wait for an empty buffer to fill with data from the backup image file. The Buffer Manipulator EDUs wait for a full buffer to write to the table space containers. |
| WaitQ | Time spent waiting for a state machine control message. When a Buffer Manipulator EDU has no more table spaces to process, the WaitQ value increases, while other Buffer Manipulators continue to process their table spaces. |
| Decompr | The amount of time that this EDU spent decompressing data. |
| LogPhase | The amount of time that this EDU spent processing recovery logs. |
| RmtStorg | The amount of time that this EDU spent communicating with a Cloud Object Storage platform. Due to internal implementation details, part of this value can sometimes be attributed to the I/O column. |
| Buffers | The total number of I/O buffers that were processed by this EDU. |
| Bytes | The quantity of data that was processed by this EDU. The column header can also be displayed as KBytes or MBytes depending on the quantity of data processed. |
| LogBytes | The quantity of recovery log data processed by this EDU. |
Next, if the restore operation is using the INCREMENTAL AUTOMATIC options, information about the
incremental backup images to be restored is
displayed.
Incremental Automatic Restore:
---------------------------------------------------------------------
currentImage: 20180716112613
currImgIndex: 2
numBckpImages: 6
numImagesRestored: 1| Field | Description |
|---|---|
| currentImage | The current backup image that is being processed. |
| currImgIndex | The zero-based index number of the current backup images that are being processed. |
| numBckpImages | The number of backup images to be restored to complete the incremental automatic restore to the TAKEN AT time that is specified in the restore command. |
| numImagesRestored | The number of backup image restore operations completed. |
Lastly, current work progress information is displayed:
Progress Monitor:
---------------------------------------------------------------------
Phase #: 2 of 3
Total work units (bytes): 18446744073709551615
Completed work units (bytes): 0
---------------------------------------------------------------------