A Administering Oracle Database on AIX

A.1.1 Controlling Buffer-Cache Paging Activity

Excessive paging activity decreases performance substantially. This can become a problem with database files created on journaled file systems (JFS and JFS2). In this situation, a large number of SGA data buffers may also have analogous file system buffers containing the most frequently referenced data. The behavior of the AIX file buffer cache manager can have a significant impact on performance. It can cause an I/O bottleneck, resulting in lower overall system throughput.

It is possible to tune buffer-cache paging activity, but you must do it carefully and infrequently. Use the /usr/sbin/vmo command to tune the AIX system parameters in the following table.

A.1.2 Tuning the AIX File Buffer Cache

The purpose of the AIX file buffer cache is to reduce disk access frequency when journaled file systems are used. If this cache is too small, then disk usage increases and potentially saturates one or more disks. If the cache is too large, then memory is wasted.

You can configure the AIX file buffer cache by adjusting the minperm and maxperm parameters. In general, if the buffer hit ratio is low (less than 90 percent), as determined by the sar -b command, then increasing the minperm parameter value may help. If maintaining a high buffer hit ratio is not critical, then decreasing the minperm parameter value increases the physical memory available. Refer to the AIX documentation for more information about increasing the size of the AIX file buffer cache.

The performance gain cannot be quantified easily, because it depends on the degree of multiprogramming and the I/O characteristics of the workload.

Tuning the minperm and maxperm Parameters

AIX provides a mechanism for you to loosely control the ratio of page frames used for files rather than those used for computational (working or program text) segments by adjusting the minperm and maxperm values according to the following guidelines:

• If the percentage of real memory occupied by file pages falls below the minperm value, then the virtual memory manager (VMM) page-replacement algorithm takes both file and computational pages, regardless of repage rates.

• If the percentage of real memory occupied by file pages rises above the maxperm value, then the VMM page-replacement algorithm takes both file and computational pages.

• If the percentage of real memory occupied by file pages is between the minperm and maxperm parameter values, then the VMM normally takes only file pages. However, if the repaging rate for file pages is higher than the repaging rate for computational pages, then the computational pages are taken as well.

Use the following algorithm to calculate the default values:

• minperm (in pages) = ((number of page frames)-1024) * 0.2

• maxperm (in pages) = ((number of page frames)-1024) * 0.8

Use the following command to change the value of the minperm parameter to 5 percent of the total number of page frames, and the value of the maxperm parameter to 20 percent of the total number of page frames:

à  # /usr/sbin/vmo -o minperm%=5 -o maxperm%=20

The default values are 20 percent and 80 percent, respectively.

To optimize for quick response when opening new database connections, adjust the minfree parameter to maintain enough free pages in the system to load the application into memory without adding additional pages to the free list. To determine the real memory size (resident set size, working set) of a process, use the following command:

$ ps v process_id

Set the minfree parameter to this value or to 8 frames, whichever is larger.

If the database files are on raw devices, or if you are using Direct I/O, then you can set the minperm and maxperm parameters to low values. For example, 5 percent and 20 percent, respectively. This is because the AIX file buffer cache is not used either for raw devices or for Direct I/O. The memory may be better used for other purposes, such as for the Oracle System Global Area.

A.1.3 Allocating Sufficient Paging Space

Inadequate paging space (swap space) usually causes the system to stop responding or show very slow response times. On AIX, you can dynamically add paging space on raw disk partitions. The amount of paging space you should configure depends on the amount of physical memory present and the paging space requirements of your applications. Use the lsps command to monitor paging space use and the vmstat command to monitor system paging activities. To increase the paging space, use the smit pgsp command.

If paging space is preallocated, then Oracle recommends that you set the paging space to a value larger than the amount of RAM. But on AIX, paging space is not allocated until required. The system uses swap space only if it runs out of real memory. If the memory is sized correctly, then there is no paging and the page space can be small. Workloads where the demand for pages does not fluctuate significantly perform well with a small paging space. Workloads likely to have peak periods of increased paging require enough paging space to handle the peak number of pages.

As a general rule, an initial setting for the paging space is half the size of RAM plus 4 GB, with an upper limit of 32 GB. Monitor the paging space use with the lsps -a command, and increase or decrease the paging space size accordingly. The metric %Used in the output of lsps -a is typically less than 25 percent on a healthy system. A properly sized deployment requires very little paging space and an excessive amount of swapping is an indication that the RAM on the system may be undersized.

A.1.4 Controlling Paging

Constant and excessive paging indicates that the real memory is over-committed. In general, you should:

• Avoid constant paging unless the system is equipped with very fast expanded storage that makes paging between memory and expanded storage much faster than Oracle Database can read and write data between the SGA and disks.

• Allocate limited memory resource to where it is most beneficial to system performance. It is sometimes a recursive process of balancing the memory resource requirements and trade-offs.

• If memory is not adequate, then build a prioritized list of memory-requiring processes and elements of the system. Assign memory to where the performance gains are the greatest. A prioritized list may look like the following:

  1. OS and RDBMS kernels

  2. User and application processes

  3. Redo log buffer

  4. PGAs and shared pool

  5. Database block buffer caches

For example, suppose that you query Oracle Database dynamic performance tables and views and find that both the shared pool and database buffer cache require more memory. Then, assigning the limited spare memory to the shared pool may be more beneficial than assigning it to the database block buffer caches.

The following AIX commands provide paging status and statistics:

• vmstat -s

• vmstat interval [repeats]

• sar -r interval [repeats]

A.1.5 Setting the Database Block Size

You can configure Oracle Database block size for better I/O throughput. On AIX, you can set the value of the DB_BLOCK_SIZE initialization parameter to between 2 KB and 32 KB, with a default of 4 KB. If Oracle Database is installed on a journaled file system, then the block size should be a multiple of the file system block size (4 KB on JFS, 16 K to 1 MB on GPFS). For databases on raw partitions, Oracle Database block size is a multiple of the operating system physical block size (512 bytes on AIX).

Oracle recommends smaller Oracle Database block sizes (2 KB or 4 KB) for online transaction processing (OLTP) or mixed workload environments and larger block sizes (8 KB, 16 KB, or 32 KB) for decision support system (DSS) workload environments.

A.1.6 Tuning the Log Archive Buffers

By increasing the LOG_BUFFER size, you may be able to improve the speed of archiving the database, particularly if transactions are long or numerous. Monitor the log file I/O activity and system throughput to determine the optimum LOG_BUFFER size. Tune the LOG_BUFFER parameter carefully to ensure that the overall performance of normal database activity does not degrade.

A.1.7 I/O Buffers and SQL*Loader

For high-speed data loading, such as using the SQL*Loader direct path option in addition to loading data in parallel, the CPU spends most of its time waiting for I/O to complete. By increasing the number of buffers, you can maximize CPU usage, and by doing this, increase overall throughput.

The number of buffers (set by the SQL*Loader BUFFERS parameter) you choose depends on the amount of available memory and how much you want to maximize CPU usage.

The performance gains depend on CPU usage and the degree of parallelism that you use when loading data.

BUFFER Parameter for the Import Utility

The BUFFER parameter for the Import utility should be set to a large value to optimize the performance of high-speed networks when they are used. For example, if you use the IBM RS/6000 Scalable POWERparallel Systems (SP) switch, then you should set the BUFFER parameter to a value of at least 1 MB.

A.2.1 AIX Logical Volume Manager

The AIX Logical Volume Manager (LVM) can stripe data across multiple disks to reduce disk contention. The primary objective of striping is to achieve high performance when reading and writing large sequential files. Effective use of the striping features in the LVM enables you to spread I/O more evenly across disks, resulting in greater overall performance.

Design a Striped Logical Volume

When you define a striped logical volume, you must specify the items listed in the following table.

Other Considerations

Performance gains from effective use of the LVM can vary greatly, depending on the LVM you use and the characteristics of the workload. For DSS workloads, you can see substantial improvement. For OLTP-type or mixed workloads, you can expect significant performance gains.

A.2.2 Using Journaled File Systems Compared to Raw Logical Volumes

Address the following considerations when deciding whether or not to use journaled file systems or raw logical volumes:

• File systems are continually being improved, as are various file system implementations. In some cases, file systems provide better I/O performance than raw devices.

• File systems require some additional configuration (AIX minservers and maxservers parameter) and add a small CPU overhead because Asynchronous I/O on file systems is serviced outside the kernel.

• Different vendors implement the file system layer in different ways to capitalize on the strengths of different disks. This makes it difficult to compare file systems across platforms.

• The introduction of more powerful LVM interfaces substantially reduces the tasks of configuring and backing up logical disks based on raw logical volumes.

• The Direct I/O and Concurrent I/O features included in AIX 5L improve file system performance to a level comparable to raw logical volumes.

If you use a journaled file system, then it is easier to manage and maintain database files than if you use raw devices. In earlier versions of AIX, file systems supported only buffered read and write and added extra contention because of imperfect inode locking. These two issues are solved by the JFS2 Concurrent I/O feature and the GPFS Direct I/O feature, enabling file systems to be used instead of raw devices, even when optimal performance is required.

File System Options

AIX 5L includes Direct I/O and Concurrent I/O support. Direct I/O and Concurrent I/O support enables database files to exist on file systems while bypassing the operating system buffer cache and removing inode locking operations that are redundant with the features provided by Oracle Database.

Where possible, Oracle recommends enabling Concurrent I/O or Direct I/O on file systems containing Oracle data files. The following table lists file systems available on AIX and the recommended setting.

Considerations for JFS and JFS2

If you are placing Oracle Database logs on a JFS2 file system, then the optimal configuration is to create the file system using the agblksize=512 option and to mount it with the cio option. This delivers logging performance within a few percentage points of the performance of a raw device.

Before Oracle Database 10g, Direct I/O and Concurrent I/O could not be enabled at the file level on JFS/JFS2. Therefore, the Oracle home directory and data files had to be placed in separate file systems for optimal performance. The Oracle home directory was placed on a file system mounted with default options, with the data files and logs on file systems mounted using the dio or cio options.

With Oracle Database 10g, you can enable Direct I/O and Concurrent I/O on JFS/JFS2 at the file level. You can do this by setting the FILESYSTEMIO_OPTIONS parameter in the server parameter file to setall or directIO. This enables Concurrent I/O on JFS2 and Direct I/O on JFS for all data file I/O. Because the directIO setting disables asynchronous I/O it should normally not be used. As a result of this 10g feature, you can place data files on the same JFS/JFS2 file system as the Oracle home directory and still use Direct I/O or Concurrent I/O for improved performance. As mentioned earlier, you should still place Oracle Database logs on a separate JFS2 file system for optimal performance.

Considerations for GPFS

If you are using GPFS, then you can use the same file system for all purposes. This includes using it for the Oracle home directory and for storing data files and logs. For optimal performance, you should use a large GPFS block size (typically, at least 512 KB). GPFS is designed for scalability, and there is no requirement to create multiple GPFS file systems as long as the amount of data fits in a single GPFS file system.

Moving from a Journaled File System to Raw Logical Volumes

To move from a journaled file system to raw devices without having to manually reload all the data, perform the following steps as the root user:

1. Create a raw device (preferably, in a BigVG) using the new raw logical volume device type (-T O), which enables putting the first Oracle block at offset zero for optimal performance:

   # mklv -T O -y new_raw_device VolumeGroup NumberOfPartitions
   
   

   
   

   Note: The raw device should be larger than the existing file. In addition, you must bear in mind the size of the new raw device to prevent wasting space.

   
   

2. Set the permissions on the raw device.

3. Use dd to convert and copy the contents of the JFS file to the new raw device as follows:

   # dd if=old_JFS_file of=new_raw_device bs=1m
   
   

4. Rename the data file.

Moving from Raw Logical Volumes to a Journaled File System

The first Oracle block on a raw logical volume is not necessarily at offset zero. However, the first Oracle block on a file system is always at offset zero. To determine the offset and locate the first block on a raw logical volume, use the $ORACLE_HOME/bin/offset command. The offset can be 4096 bytes or 128 KB on AIX logical volumes or zero on AIX logical volumes created with the mklv -T O option.

When you have determined the offset, you can copy over data from a raw logical volume to a file system using the dd command and skipping the offset. The following example assumes an offset of 4096 bytes:

# dd if=old_raw_device bs=4k skip=1|dd of=new_file bs=256

You can instruct Oracle Database to use a number of blocks smaller than the maximum capacity of a raw logical volume. If you do this, then you must add a count clause to ensure that only data that contains Oracle blocks is copied. The following example assumes an offset of 4096 bytes, an Oracle block size of 8 KB, and 150000 blocks:

# dd if=old_raw_device bs=4k skip=1|dd bs=8k count=150000|dd of=new_file bs=256k

A.2.3 Using Asynchronous I/O

Oracle Database takes full advantage of asynchronous I/O (AIO) provided by AIX, resulting in faster database access.

AIX 5L supports asynchronous I/O (AIO) for database files created both on file system partitions and on raw devices. AIO on raw devices is implemented fully into the AIX kernel, and does not require database processes to service the AIO requests. When using AIO on file systems, the kernel database processes (aioserver) control each request from the time a request is taken off the queue to the time it is completed. The number of aioserver servers determines the number of AIO requests that can be processed in the system concurrently. So, it is important to tune the number of aioserver processes when using file systems to store Oracle Database data files.

Use one of the following commands to set the number of servers. This applies only when using asynchronous I/O on file systems rather than raw devices:

• smit aio

• chdev -l aio0 -a maxservers=' m ' -a minservers='n'

  
  

  See Also: The System Management Interface Tool (SMIT) online Help for more information about SMIT The man pages for more information about the smit aio and chdev commands

  
  

Set the minimum value to the number of servers to be started when the system is started. Set the maximum value to the number of servers that can be started in response to a large number of concurrent requests. These parameters apply to file systems only. They do not apply to raw devices.

The default value for the minimum number of servers is 1. The default value for the maximum number of servers is 10. These values are usually too low to run Oracle Database on large systems with 4 CPUs or more, if you are not using kernelized AIO. Oracle recommends that you set the parameters to the values listed in the following table.

$ lsattr -E -l hdiskxx

If the value of the maxservers or maxreqs parameter is set too low, then the following warning messages are repeatedly displayed:

Warning: lio_listio returned EAGAINPerformance degradation may be seen.

You can avoid these errors by increasing the value of the maxservers parameter. To display the number of AIO servers running, enter the following commands as the root user:

# pstat -a | grep -c aios
# ps -k | grep aioserver

Check the number of active AIO servers periodically, and change the values of the minservers and maxservers parameters if required. The changes take place when the system is restarted.

A.2.4 I/O Slaves

I/O Slaves are specialized Oracle processes that perform only I/O. They are rarely used on AIX, because asynchronous I/O is the default and recommended way for Oracle to perform I/O operations on AIX. I/O Slaves are allocated from shared memory buffers. I/O Slaves use the initialization parameters listed in the following table.

Generally, you do not adjust the parameters in the preceding table. However, on large workloads, the database writer may become a bottleneck. If it does, then increase the value of DB_WRITER_PROCESSES. As a general rule, do not increase the number of database writer processes above one for each pair of CPUs in the system or partition.

There are times when you must turn off asynchronous I/O. For example, if instructed to do so by Oracle Support for debugging. You can use the DISK_ASYNCH_IO and TAPE_ASYNCH_IO parameters to switch off asynchronous I/O for disk or tape devices. Because the number of I/O slaves for each process type defaults to zero, by default, no I/O Slaves are deployed.

Set the DBWR_IO_SLAVES parameter to greater than 0 only if the DISK_ASYNCH_IO or TAPE_ASYNCH_IO parameter is set to false. Otherwise, the database writer process (DBWR) becomes a bottleneck. In this case, the optimal value on AIX for the DBWR_IO_SLAVES parameter is 4.

A.2.5 Using the DB_FILE_MULTIBLOCK_READ_COUNT Parameter

When using Direct I/O or Concurrent I/O with Oracle Database 10g, the AIX file system does not perform any read-ahead on sequential scans. For this reason the DB_FILE_MULTIBLOCK_READ_COUNT value in the server parameter file should be increased when Direct I/O or Concurrent I/O is enabled on Oracle data files. The read ahead is performed by Oracle Database as specified by the DB_FILE_MULTIBLOCK_READ_COUNT initialization parameter.

Setting a large value for the DB_FILE_MULTIBLOCK_READ_COUNT initialization parameter usually yields better I/O throughput on sequential scans. On AIX, this parameter ranges from 1 to 512, but using a value higher than 16 usually does not provide additional performance gain.

Set this parameter so that its value when multiplied by the value of the DB_BLOCK_SIZE parameter produces a number larger than the LVM stripe size. Such a setting causes more disks to be used.

A.2.6 Using Write Behind

The write behind feature enables the operating system to group write I/Os together, up to the size of a partition. You can improve performance by doing this, because the number of I/O operations is reduced. The file system divides each file into 16 KB partitions to increase write performance, limit the number of dirty pages in memory, and minimize disk fragmentation. The pages of a particular partition are not written to disk until the program writes the first byte of the next 16 KB partition. To set the size of the buffer for write behind to eight 16 KB partitions, enter the following command:

à  # /usr/sbin/vmo -o numclust=8

To disable write behind, enter the following command:

à  # /usr/sbin/vmo -o numclust=0 

A.2.7 Tuning Sequential Read Ahead

The VMM anticipates the need for pages of a sequential file. It observes the pattern in which a process accesses a file. When the process accesses two consecutive pages of the file, the VMM assumes that the program continues to access the file sequentially, and schedules additional sequential reads of the file. These reads overlap the program processing and make data available to the program faster. The following VMM thresholds, implemented as kernel parameters, determine the number of pages it reads ahead:

• minpgahead

  This parameter stores the number of pages read ahead when the VMM first detects the sequential access pattern.

• maxpgahead

  This parameter stores the maximum number of pages that VMM reads ahead in a sequential file.

Set the minpgahead and maxpgahead parameters to appropriate values for an application. The default values are 2 and 8 respectively. Use the /usr/sbin/vmo command to change these values. You can use higher values for the maxpgahead parameter in systems where the sequential performance of striped logical volumes is of paramount importance. To set the minpgahead parameter to 32 pages and the maxpgahead parameter to 64 pages, run the following command as the root user:

-o minpgahead=32 -o maxpgahead=64

Set both the minpgahead and maxpgahead parameters to a power of two. For example, 2, 4, 8, . . . 512, 1042, . . . and so on.

A.2.8 Tuning Disk I/O Pacing

Disk I/O pacing is an AIX mechanism that enables the system administrator to limit the number of pending I/O requests to a file. This prevents disk I/O intensive processes from saturating the CPU. Therefore, the response time of interactive and CPU-intensive processes does not deteriorate.

You can achieve disk I/O pacing by adjusting two system parameters: the high-water mark and the low-water mark. When a process writes to a file that already has a pending high-water mark I/O request, the process is put to sleep. The process wakes up when the number of outstanding I/O requests falls below or equals the low-water mark.

You can use the smit command to change the high and low-water marks. Determine the water marks through trial-and-error. Use caution when setting the water marks, because they affect performance. Tuning the high and low-water marks has less effect on disk I/O larger than 4 KB.

You can determine disk I/O saturation by analyzing the result of iostat, in particular, the percentage of iowait and tm_act. A high iowait percentage combined with high tm_act percentages on specific disks is an indication of disk saturation. Note that a high iowait alone is not necessarily an indication of an I/O bottleneck.

A.2.9 Resilvering with Oracle Database

If you disable mirror write consistency (MWC) for an Oracle data file allocated on a raw logical volume (LV), then the Oracle Database crash recovery process uses resilvering to recover after a system failure. This resilvering process prevents database inconsistencies or corruption.

During crash recovery, if a data file is allocated on a logical volume with more than one copy, then the resilvering process performs a checksum on the data blocks of all the copies. It then performs one of the following:

• If the data blocks in a copy have valid checksums, then the resilvering process uses that copy to update the copies that have invalid checksums.

• If all copies have blocks with invalid checksums, then the resilvering process rebuilds the blocks using information from the redo log file. It then writes the data file to the logical volume and updates all the copies.

On AIX, the resilvering process works only for data files allocated on raw logical volumes for which MWC is disabled. Resilvering is not required for data files on mirrored logical volumes with MWC enabled, because MWC ensures that all copies are synchronized.

If the system crashes while you are upgrading an earlier release of Oracle Database that used data files on logical volumes for which MWC was disabled, then run the syncvg command to synchronize the mirrored LV before starting Oracle Database. If you do not synchronize the mirrored LV before starting the database, then Oracle Database may read incorrect data from an LV copy.

A.2.10 Backing Up Raw Devices

Oracle recommends that you use RMAN to back up raw devices. If you do use the dd command to back up raw devices, then follow the instructions described in this section.

The offset of the first Oracle block on a raw device may be 0, 4K, or 128K depending on the device type. You can use the offset command to determine the proper offset.

When creating a logical volume, Oracle recommends using an offset of zero, which is possible if you use -T O option. However, existing raw logical volumes created with earlier versions of Oracle Database typically have a nonzero offset. The following example shows how to back up and restore a raw device whose first Oracle block is at offset 4K:

$ dd if=/dev/raw_device of=/dev/rmt0.1 bs=256k

To restore the raw device from tape, enter commands similar to the following:

$ dd if=/dev/rmt0.1 of=/dev/raw_device count=63 seek=1 skip=1 bs=4k
$ mt -f /dev/rmt0.1 bsf 1
$ dd if=/dev/rmt0.1 of=/dev/raw_device seek=1 skip=1 bs=256k

A.3.1 Changing Process Running Time Slice

The default value for the run-time slice of the AIX RR dispatcher is ten milliseconds (msec). Use the schedo command to change the time slice. A longer time slice causes a lower context switch rate if the average voluntary switch rate of the applications is lower. As a result, fewer CPU cycles are spent on context-switching for a process and the system throughput should improve.

However, a longer run-time slice can deteriorate response time, especially on a uniprocessor system. The default run-time slice is usually acceptable for most applications. When the run queue is high and most of the applications and Oracle shadow processes are capable of running a much longer duration, you may want to increase the time slice by entering the following command:

# /usr/sbin/schedo -t n

In the preceding command, setting n to 0 results in a slice of 10 msec, choosing a value of 1 results in a slice of 20 msec, choosing a value of 2 results in a slice of 30 msec, and so on.

A.3.2 Using Processor Binding on SMP Systems

Binding certain processes to a processor can improve performance substantially on an SMP system. Processor binding is available and fully functional on AIX 5L.

However, starting with AIX 5L version 5.2, specific improvements in the AIX scheduler enables Oracle Database processes to be scheduled optimally without processor binding. Therefore, Oracle no longer recommends binding processes to processors when running on AIX 5L version 5.2 or later.