A closer look at ASM rebalance, Part III: Disks have been added and dropped (at the same time)

This article is the third Part of the “A closer look at ASM rebalance” series:

1. Part I: Disks have been added.
2. Part II: Disks have been dropped.
3. Part III: Disks have been added and dropped (at the same time).

If you are not familiar with ASM rebalance I would suggest first to read those 2 blog posts written by Bane Radulovic:

• Rebalancing act
• When will my rebalance complete

In this part III I want to visualize the rebalance operation (with 3 power values: 2,6 and 11) after disks have been added and dropped (at the same time).

To do so, on a 2 nodes Extended Rac Cluster (11.2.0.4), I added 2 disks and dropped 2 disks (with a single command) into the DATA diskgroup (created with an ASM Allocation Unit of 4MB) and launched (connected on +ASM1):

1. alter diskgroup DATA rebalance power 2; (At 02:11 PM).
2. alter diskgroup DATA rebalance power 6; (At 02:24 PM).
3. alter diskgroup DATA rebalance power 11; (At 02:34 PM).

And then I waited until it finished (means v$asm_operation returns no rows for the DATA diskgroup).

Note that 2) and 3) interrupted the rebalance in progress and launched a new one with a new power.

During this amount of time I collected the ASM performance metrics that way for the DATA diskgroup only.

I’ll present the results with Tableau (For each Graph I’ll keep the “columns”, “rows” and “marks” shelf into the print screen so that you can reproduce).

Note: There is no database activity on the Host where the rebalance has been launched.

Here are the results:

First let’s verify that the whole rebalance activity has been done on the +ASM1 instance (As I launched the rebalance operations from it).

We can see:

1. That all Read and Write rebalance activity has been done on +ASM1 .
2. That the read throughput is very close to the write throughput on +ASM1.
3. The impact of the power values (2,6 and 11) on the throughput.

Now I would like to compare the behavior of 3 Sets of Disks: The disks that have been dropped, the disks that have been added and the other existing disks into the DATA diskgroup.

To do so, let’s create in Tableau 3 groups:

Let’s call it “3 Groups”

So that now we are able to display the ASM metrics for those 3 sets of disks.

I will filter the metrics on ASM1 only (to avoid any “little parasites” coming from ASM2).

Let’s visualize the Reads/s and Writes/s metrics:

We can see that during the 3 rebalances:

1. No writes on the dropped disks.
2. No reads on the new disks.
3. Number of Reads/s increasing on the dropped disks depending of the power values.
4. Number of Writes/s increasing on the new disks depending of the power values.
5. Reads/s and Writes/s both increasing on the other disks depending of the power values.
6. As of 03:06 PM, no activity on the dropped and new disks while there is still activity on the other disks.

• Are 1, 2, 3, 4 and 5 surprising? No.
• What happened for 6? I’ll answer later on.

Let’s visualize the Kby Read/s and Kby Write/s metrics:

We can see that during the 3 rebalances:

1. No Kby Write/s on the dropped disks.
2. No Kby Read/s on the new disks.
3. Number of Kby Read/s increasing on the dropped disks depending of the power values.
4. Number of Kby Write/s increasing on the new disks depending of the power values.
5. Kby Read/s and Kby Write/s both increasing on the other disks depending of the power values.
6. Kby Read/s and Kby Write/s are very close on the other disks (It was not the case into the Part I).
7. As of 03:06 PM, no activity on the dropped and new disks while there is still activity on the other disks.

• Are 1, 2, 3, 4, 5 and 6 surprising? No.
• What happened for 7? I’ll answer later on.

Let’s visualize the Average By/Read and Average By/Write metrics:

Important remark regarding the averages computation/display: The By/Read and By/Write measures depend on the number of reads. So the averages have to be calculated using Weighted Averages.

Let’s create the calculated field in Tableau for the By/Read Weighted Average:

The same has to be done for the By/Write Weighted Average.

Let’s see the result:

We can see:

1. The Avg By/Read on the dropped disks is about the same (about 1MB) whatever the power value is.
2. The Avg By/Write on the new disks is about the same (about 1MB) whatever the power value is.
3. The Avg By/Read and Avg By/Write on the other disks is about the same (about 1MB) whatever the power value is.

• Are 1,2 and 3 surprising? No for the behaviour,Yes (at least for me) for the 1MB value as the ASM allocation unit is 4MB.

Now that we have seen all those metrics, we can ask:

Q1: So what the hell happened at 03:06 pm?

Let’s check the alert_+ASM1.log file at that time:

Mon Aug 25 15:06:13 2014
NOTE: membership refresh pending for group 4/0x1e089b59 (DATA)
GMON querying group 4 at 396 for pid 18, osid 67864
GMON querying group 4 at 397 for pid 18, osid 67864
NOTE: Disk DATA_0006 in mode 0x0 marked for de-assignment
NOTE: Disk DATA_0007 in mode 0x0 marked for de-assignment
SUCCESS: refreshed membership for 4/0x1e089b59 (DATA)
NOTE: Attempting voting file refresh on diskgroup DATA
NOTE: Refresh completed on diskgroup DATA. No voting file found.
Mon Aug 25 15:07:16 2014
NOTE: stopping process ARB0
SUCCESS: rebalance completed for group 4/0x1e089b59 (DATA)

We can see that the ASM rebalance started the compacting phase (See Bane Radulovic’s blog post for more details about the ASM rebalances phases).

Q2: The ASM Allocation Unit size is 4MB and the Avg By/Read is stucked to 1MB,why?

I don’t have the answer yet, it will be the subject of another post.

Two remarks before to conclude:

1. The ASM rebalance activity is not recorded into the v$asm_disk_iostat view. It is recorded into the v$asm_disk_stat view. So, if you are using the asm_metrics utility, you have to change the asm_feature_version variable to a value > your ASM instance version.
2. I tested with compatible.asm set to 10.1 and 11.2.0.2 and observed the same behaviour for all those metrics.

Conclusion of Part III:

• Nothing surprising except (at least for me) that the Avg By/Read is stucked to 1MB (While the allocation unit is 4MB).
• We visualized that the compacting phase of the rebalance operation generates much more activity on the other disks compare to near zero activity on the dropped and new disks.
• I’ll update this post with ASM 12c results as soon as I can (if something new needs to be told).

Advertisement

A closer look at ASM rebalance, Part II: Disks have been dropped

This article is the second Part of the “A closer look at ASM rebalance” series:

1. Part I: Disks have been added.
2. Part II: Disks have been dropped.
3. Part III: Disks have been added and dropped (at the same time).

If you are not familiar with ASM rebalance I would suggest first to read those 2 blog posts written by Bane Radulovic:

• Rebalancing act
• When will my rebalance complete

In this part II I want to visualize the rebalance operation (with 3 power values: 2,6 and 11) after disks have been dropped.

To do so, on a 2 nodes Extended Rac Cluster (11.2.0.4), I dropped 2 disks into the DATA diskgroup (created with an ASM Allocation Unit of 4MB) and launched (connected on +ASM1):

1. alter diskgroup DATA rebalance power 2; (At 09:09 AM).
2. alter diskgroup DATA rebalance power 6; (At 09:19 AM).
3. alter diskgroup DATA rebalance power 11; (At 09:29 AM).

And then I waited until it finished (means v$asm_operation returns no rows for the DATA diskgroup).

Note that 2) and 3) interrupted the rebalance in progress and launched a new one with a new power.

During this amount of time I collected the ASM performance metrics that way for the DATA diskgroup only.

I’ll present the results with Tableau (For each Graph I’ll keep the “columns”, “rows” and “marks” shelf into the print screen so that you can reproduce).

Note: There is no database activity on the Host where the rebalance has been launched.

Here are the results:

First let’s verify that the whole rebalance activity has been done on the +ASM1 instance (As I launched the rebalance operations from it).

We can see:

1. That all Read and Write rebalance activity has been done on +ASM1 .
2. That the read throughput is very close to the write throughput on +ASM1.
3. The impact of the power values (2,6 and 11) on the throughput.

Now I would like to compare the behavior of 2 Sets of Disks: The disks that have been dropped and the disks that will remain into the DATA diskgroup.

To do so, let’s create in Tableau a SET that contains the 2 dropped disks.

Let’s call it “Dropped Disks”

So that now we are able to display the ASM metrics IN this set (the 2 dropped disks) and OUT this set (the disks that will remain into the DATA diskgroup).

I will filter the metrics on ASM1 only (to avoid any “little parasites” coming from ASM2).

Let’s visualize the Reads/s and Writes/s metrics:

We can see that during the 3 rebalances:

1. No writes on the dropped disks.
2. Number of Reads/s increasing on the dropped disks depending of the power values.
3. Reads/s and Writes/s both increasing on the remaining disks depending of the power values.

• Are 1, 2 and 3 surprising? No.

Let’s visualize the Kby Read/s and Kby Write/s metrics:

We can see that during the 3 rebalances:

1. No Kby Write/s on the dropped disks.
2. Number of Kby Read/s increasing on the dropped disks depending of the power values.
3. Kby Read/s and Kby Write/s both increasing on the remaining disks depending of the power values.

Are 1, 2 and 3 surprising? No.

Let’s visualize the Average By/Read and Average By/Write metrics:

Important remark regarding the averages computation/display: The By/Read and By/Write measures depend on the number of reads. So the averages have to be calculated using Weighted Averages.

Let’s create the calculated field in Tableau for the By/Read Weighted Average:

The same has to be done for the By/Write Weighted Average.

Let’s see the result:

We can see:

1. The Avg By/Read on the dropped disks is about the same (about 1MB) whatever the power value is.
2. The Avg By/Read and Avg By/Write on the remaining disks is about the same (about 1MB) whatever the power value is.

• Are 1 and 2 surprising? No for the behaviour, Yes (at least for me) for the 1MB value as the ASM allocation unit is 4MB.

Now that we have seen all those metrics, we can ask:

Q1: The ASM Allocation Unit size is 4MB and the Avg By/Read is stucked to 1MB,why?

I don’t have the answer yet, it will be the subject of another post.

Two remarks before to conclude:

1. The ASM rebalance activity is not recorded into the v$asm_disk_iostat view. It is recorded into the v$asm_disk_stat view. So, if you are using the asm_metrics utility, you have to change the asm_feature_version variable to a value > your ASM instance version.
2. I tested with compatible.asm set to 10.1 and 11.2.0.2 and observed the same behaviour for all those metrics.

Conclusion of Part II:

• Nothing surprising except (at least for me) that the Avg By/Read is stucked to 1MB (While the allocation unit is 4MB).
• I’ll update this post with ASM 12c results as soon as I can (if something new needs to be told).

A closer look at ASM rebalance, Part I: Disks have been added

This article is the first Part of the “A closer look at ASM rebalance” series:

1. Part I: Disks have been added.
2. Part II: Disks have been dropped.
3. Part III: Disks have been added and dropped (at the same time).

If you are not familiar with ASM rebalance I would suggest first to read those 2 blog posts written by Bane Radulovic:

• Rebalancing act
• When will my rebalance complete

In this part I want to visualize the rebalance operation (with 3 power values: 2,6 and 11) after disks have been added (no dropped disks yet: It will be for the parts II and III).

To do so, on a 2 nodes Extended Rac Cluster (11.2.0.4), I added 2 disks into the DATA diskgroup (created with an ASM Allocation Unit of 4MB) and launched (connected on +ASM1):

1. alter diskgroup DATA rebalance power 2; (At 11:55 AM).
2. alter diskgroup DATA rebalance power 6; (At 12:05 PM).
3. alter diskgroup DATA rebalance power 11; (At 12:15 PM).

And then I waited until it finished (means v$asm_operation returns no rows for the DATA diskgroup).

Note that 2) and 3) interrupted the rebalance in progress and launched a new one with a new power.

During this amount of time I collected the ASM performance metrics that way for the DATA diskgroup only.

I’ll present the results with Tableau (For each Graph I’ll keep the “columns”, “rows” and “marks” shelf into the print screen so that you can reproduce).

Note: There is no database activity on the Host where the rebalance has been launched.

Here are the results:

First let’s verify that the whole rebalance activity has been done on the +ASM1 instance (As I launched the rebalance operations from it).

We can see:

1. That all Read and Write rebalance activity has been done on +ASM1 .
2. That the read throughput is very close to the write throughput on +ASM1.
3. The impact of the power values (2,6 and 11) on the throughput.

Now I would like to compare the behavior of 2 Sets of Disks: The disks added and the disks that are already part of the DATA diskgroup.

To do so, let’s create in Tableau a SET that contains the 2 new disks.

Let’s call it “New Disks”

So that now we are able to display the ASM metrics IN this set (the 2 new disks) and OUT this set (the already existing disks means already part of the DATA diskgroup).

I will filter the metrics on ASM1 only (to avoid any “little parasites” coming from ASM2).

Let’s visualize the Reads/s and Writes/s metrics:

We can see that during the 3 rebalances:

1. No Reads on the new disks (at least until about 12:40 pm).
2. Number of Writes/s increasing on the new disks depending of the power values.
3. Reads/s and Writes/s both increasing on the already existing disks depending of the power values.
4. As of 12.40 pm, activity on the existing disks while near zero activity on the new ones.
5. As of 12.40 pm number of Writes/s >= Reads/s on the existing disks (while it was the opposite before).

• Are 1, 2 and 3 surprising? No.
• What happened for 4 and 5? I’ll answer later on.

Let’s visualize the Kby Read/s and Kby Write/s metrics:

We can see that during the 3 rebalances:

1. No Kby Read/s on the new disks.
2. Number of Kby Write/s increasing on the new disks depending of the power values.
3. Kby Read/s and Kby Write/s both increasing on the existing disks depending of the power values.
4. As of 12.40 pm, activity on the existing disks while no activity on the new ones.
5. As of 12.40 pm same amount of Kby Read/s and Kby Write/s on the existing disks (while it was not the case before).

• Are 1, 2 and 3 surprising? No.
• What happened for 4 and 5? I’ll answer later on.

Let’s visualize the Average By/Read and Average By/Write metrics:

Important remark regarding the averages computation/display: The By/Read and By/Write measures depend on the number of reads. So the averages have to be calculated using Weighted Averages.

Let’s create the calculated field in Tableau for the By/Read Weighted Average:

The same has to be done for the By/Write Weighted Average.

Let’s see the result:

We can see:

1. The Avg By/Write on the new disks is about the same (about 1MB) whatever the power value is (before 12:40 pm).
2. The Avg By/Write tends to increase with the power on the already existing disks.
3. The Avg By/Read on the existing disks is about the same (about 1MB) whatever the power value is.

• Is 1 surprising? No.
• Is 2 surprising? Yes (at least for me).
• Is 3 surprising? No.

Now that we have seen all those metrics, we can ask:

Q1: So what the hell happened at 12:40 pm?

Let’s check the alert_+ASM1.log file at that time:

Mon Aug 25 12:15:44 2014
ARB0 started with pid=33, OS id=1187132
NOTE: assigning ARB0 to group 4/0x1e089b59 (DATA) with 11 parallel I/Os
Mon Aug 25 12:15:47 2014
NOTE: Attempting voting file refresh on diskgroup DATA
NOTE: Refresh completed on diskgroup DATA. No voting file found.
cellip.ora not found.
Mon Aug 25 12:39:52 2014
NOTE: GroupBlock outside rolling migration privileged region
NOTE: requesting all-instance membership refresh for group=4
Mon Aug 25 12:40:03 2014
GMON updating for reconfiguration, group 4 at 372 for pid 35, osid 1225810
NOTE: group DATA: updated PST location: disk 0014 (PST copy 0)
NOTE: group DATA: updated PST location: disk 0015 (PST copy 1)
Mon Aug 25 12:40:03 2014
NOTE: group 4 PST updated.
Mon Aug 25 12:40:03 2014
NOTE: membership refresh pending for group 4/0x1e089b59 (DATA)
GMON querying group 4 at 373 for pid 18, osid 67864
SUCCESS: refreshed membership for 4/0x1e089b59 (DATA)
NOTE: Attempting voting file refresh on diskgroup DATA
NOTE: Refresh completed on diskgroup DATA. No voting file found.
Mon Aug 25 12:45:24 2014
NOTE: F1X0 copy 2 relocating from 18:44668 to 18:20099 for diskgroup 4 (DATA)
Mon Aug 25 12:53:49 2014
NOTE: stopping process ARB0
SUCCESS: rebalance completed for group 4/0x1e089b59 (DATA)

We can see that the ASM rebalance started the compacting phase (See Bane Radulovic’s blog post for more details about the ASM rebalances phases).

Q2: The ASM Allocation Unit size is 4MB and the Avg By/Read is stucked to 1MB,why?

I guess this is somehow related to the max_sectors_kb and max_hw_sectors_kb SYSFS parameters. It will be the subject of another post.

Two remarks before to conclude:

1. The ASM rebalance activity is not recorded into the v$asm_disk_iostat view. It is recorded into the v$asm_disk_stat view. So, if you are using the asm_metrics utility, you have to change the asm_feature_version variable to a value > your ASM instance version.
2. I tested with compatible.asm set to 10.1 and 11.2.0.2 and observed the same behavior for all those metrics.

Conclusion of Part I:

• We visualized that the compacting phase of the rebalance operation generates much more activity on the existing disks compare to near zero activity on the new disks.
• We visualized that the compacting phase of the rebalance operation generates the same amount of Kby Read/s and Kby Write/s on the existing disks (while it was not the case before).
• We visualized that during the compacting phase the number of Writes/s >= Reads/s on the existing disks (while it was the opposite before).
• We visualized that the Avg By/Read does not exceed 1MB on the existing disks (while the ASM allocation Unit has been set to 4MB on my diskgroup).
• We visualized that the Avg By/Write tends to increase with the power on the already existing disks (quite surprising to me).
• I’ll update this post with ASM 12c results as soon as I can (if something new needs to be told).

Are ASM rebalance and preferred read working together?

Introduction:

If I add disks into a diskgroup, then during the rebalance operation, ASM needs to read the data coming from the disks already part of the diskgroup to rebalance them on all the disks (including the new ones).

Question:

If I add 2 disks (one into each failgroup) is the preferred feature took into account for the rebalance process? (“for” means “for the reads generated by the rebalance operation”).

Let’s see:

• Set the preferred read on +ASM1 (so that +ASM1 “prefers” to read from the “HOST31” failgroup):

SQL> alter system set asm_preferred_read_failure_groups='DATA.HOST31';

System altered.

• Add 2 disks (one into each failgroup) into the DATA diskgroup (connected on +ASM1):

SQL> alter diskgroup DATA add failgroup HOST31 disk '/dev/san/HOST31CA8D0D' failgroup HOST32 disk '/dev/san/HOST32CA8D0D';

Diskgroup altered.

• Check that the ASM compatibility is high enough (>=11.1) to use the preferred read feature:

SQL> select COMPATIBILITY from v$asm_diskgroup where NAME='DATA';

COMPATIBILITY
------------------------------------------------------------
11.2.0.2.0

• Launch the rebalance:

SQL> alter diskgroup DATA rebalance power 2;

Diskgroup altered.

Now, let’s collect the ASM metrics that way and visualize the result with Tableau.

Note: During the rebalance near zero database activity occurred so that near 100% of the activity is coming from the rebalance process.

Result:

As you can see:

1. The +ASM1 instance reads from the HOST31 and the HOST32 failgroups: It did not take into account the preferred read.
2. I changed the power during the rebalance just for the fun 😉

Remark:

It has been tested on a 11.2.0.4 extended RAC (Still need to test on 12c).

Conclusion:

• The ASM preferred read feature is not took into account for the rebalance process.
• I guess it is still took into account for the reads coming from the databases during the rebalance process.

Simulate and Visualize the impact of the ASM preferred feature on the read IOPS and throughput

Suppose that you decided to put the ASM preferred feature in place because you observed that the read latency is too high on the farthest disk array (You can find how you can lead to this conclusion with the use case 3 into this post).

So, you want to enable the ASM preferred read feature so that:

1. The +ASM1 instance “prefers” to read from the “WIN” failgroup.
2. The +ASM2 instance “prefers” to read from the “JMO” failgroup.

But doing so may have an impact on the number of read IOPS and the throughput repartition per host/disk array because:

1. The “previous” ASM1 to JMO reads will now be done on the “WIN” array.
2. The “previous” ASM2 to WIN reads will now be done on the “JMO” array.

Of course, the total number of read operations and throughput will not change, but the repartition across the failgroup (disk array) may change with the ASM preferred read feature in place.

Question:

• Is the architecture able to deal with this new read repartition?

To answer this question I will:

1. Collect the ASM metrics during a certain amount of time (without the ASM preferred read in place) and produce a csv file as described here.
2. Visualize the ASM metrics with Tableau and simulate the impact of the preferred read feature on the read IOPS and the throughput repartition.

Once the csv file is ready (means you collected a representative workload), let’s check what the current workload is (Without the ASM preferred read in place).

For the Kby Read/s measure:

We can visualize it that way with Tableau (I keep the “columns”, “rows” and “marks” shelf into the print screen so that you can reproduce).

For the Reads/s measure:

We can see the read IOPS and the throughput repartition by failgroup and ASM instances. We can see that the read IOPS and the throughput are equally distributed over the Failgroups (It is the expected behaviour without the ASM preferred read in place).

Now, what If we implement the ASM preferred feature? What would be the impact on the read IOPS and the throughput repartition?

To simulate and visualize the impact, let’s create this “New FG for Read operations” calculated field:

Basically it simulates the ASM preferred Read in place by assigning the failgroup per ASM instances.

Now, let’s simulate and visualize the impact of the ASM preferred read feature (should it be implemented) using the same csv file and this calculated field as dimension.

For the Kby Read/s measure:

Note that the throughput repartition would not be the same and that the peak are higher (> 200 Mo/s compare to about 130 Mo/s without the ASM preferred read).

For the Reads/s measure:

Note that the read IOPS repartition would not be the same and that the peak on the WIN failgroup is higher (about 8000 Reads/s compare to about 5000 Reads/s without the ASM preferred read).

Now you can check (with your Systems and Storage administrators) if your current architecture would be able to deal with this new repartition.

Remarks:

• ASM is not performing any reads for the database, it records metrics for the database instances that it is servicing.

• I would not suggest to use the Flex ASM feature with the ASM preferred read because the preferred read feature is broken with Flex ASM 12c (12.1) in place.

Conclusion:

We have been able to simulate and visualize the impact of the ASM preferred read feature on the read IOPS and the throughput repartition without actually implementing it.

ASM performance metrics visualization: Use cases

Now that I can graph ASM performance metrics, let’s see some use cases.

To display the ASM metrics I’ll use the csv file generated by the csv_asm_metrics utility and Tableau for the visualization. Of course you could use the visualization tool of your choice.

Use case 1: Display the TOP IO consumers

You consolidated several databases on the same machine and you want to visualize which database is generating most of the IO throughput for Reads. You can visualize this that way with Tableau (I keep the “columns”, “rows” and “marks” shelf into the print screen so that you can reproduce with Tableau) :

I can see that one of my databases is generating most of the throughput.

Should you use RAC, you could split those metrics per ASM instances as well:

I can see that most of the activity is recorded on ASM2, which makes sense as my RAC services are configured as preferred/available (Active/Passive configuration) and started on the *_2 database instances (linked to ASM2).

Use case 2: I want to see the Read IO distributions by Failgroups

You can visualize this that way with Tableau (I keep the “columns”, “rows” and “marks” shelf into the print screen so that you can reproduce with Tableau):

We can see that the IOs are equally distributed over the Failgroups. It is the expected behaviour as I am not using the ASM Preferred Read feature.

Use case 3: Should I use the ASM Preferred Read feature on my extended RAC?

Suppose the host hosting the ASM1 instance is close to the disk array on which the “WIN” failgroup has been created. The same way, the host hosting the ASM2 instance is close to the disk array on which the “JMO” failgroup has been created. Let’s see the Read IO latency between the ASM instances and the failgroups.

You can visualize this that way with Tableau (I keep the “columns”, “rows” and “marks” shelf into the print screen so that you can reproduce with Tableau):

As you can see the ASM1 instance reports faster reads on the “WIN” failgroup and ASM2 reports faster reads on the “JMO” failgroup which makes sense according to our configuration. I can also check if the reads performance are good enough when the Reads are done on the farthest disk array (ASM1 “reading” on the JMO failgroup and ASM2 “reading” on the WIN failgroup) and then decide if the ASM Preferred Read feature needs to be implemented.

Use case 4: Simulate and Visualize the impact of the ASM preferred feature on the read IOPS and throughput (See this blog post).

Use case 5: Are ASM rebalance and preferred read working together? (See this blog post)

Use case 6: A closer look at ASM rebalance, Part I: Disks have been added. (See this blog post)

Use case 7: A closer look at ASM rebalance, Part II: Disks have been dropped. (See this blog post)

Use case 8: A closer look at ASM rebalance, Part III: Disks have been added and dropped (at the same time). (See this blog post)

Remarks:

• ASM is not performing any reads for the database, it records metrics for the database instances that it is servicing.

• I am not using the Flex ASM feature for the previous use cases. With the Flex ASM feature in place the interpretations done into the use case 3 would not have been possible (as ASM1 could serve database instance located near to ASM2 and vice versa). Furthermore the preferred read feature is broken with Flex ASM 12c (12.1) in place.

• You can imagine a lot of use cases thanks to the measures collected (Reads/s, Kby Read/s, ms/Read, By/Read, Writes/s, Kby Write/s, ms/Write, By/Write) and all those dimensions (Snap Time, INST, DBINST, DG, FG, DSK).

• Don’t forget that If you don’t picture it, you may not understand it.

You can download the asm_metrics and the csv_asm_metrics utilities from this repository.

Graphing ASM performance metrics

ASM metrics are a goldmine, they provide a lot of informations. As you may know, the asm_metrics utility extracts them in real-time.

But sometimes it is not easy to understand the values without the help of a graph. Look at this example: If I cant’ picture it, I can’t understand it.

So depending on your needs, depending on what you are looking for with the ASM metrics: A picture may help.

So let’s graph the output of the asm_metrics utility: For this I created the csv_asm_metrics utility to produce a csv file from the output of the asm_metrics utility.

Once you get the csv file you can graph the metrics with your favourite visualization tool (I’ll use Tableau as an example).

First you have to launch the asm_metrics utility that way (To ensure that all the fields are displayed):

• -show=inst,dbinst,fg,dg,dsk for ASM >= 11g
• -show=inst,fg,dg,dsk for ASM < 11g

and redirect the output to a text file:

./asm_metrics.pl -show=inst,dbinst,fg,dg,dsk > asm_metrics.txt

Remark: You can use the -interval parameter to collect data with an interval greater than one second (the default interval), as it could produce a huge output file.

The output file looks like:

............................
Collecting 1 sec....
............................

......... SNAP TAKEN AT ...................

13:48:54                                                                              Kby       Avg       AvgBy/               Kby       Avg        AvgBy/ 
13:48:54   INST     DBINST        DG            FG           DSK            Reads/s   Read/s    ms/Read   Read      Writes/s   Write/s   ms/Write   Write  
13:48:54   ------   -----------   -----------   ----------   ----------     -------   -------   -------   ------    ------     -------   --------   ------ 
13:48:54   +ASM1                                                            6731      54224     1.4       8249      42         579       3.0        14117  
13:48:54   +ASM1    BDT10_1                                                 2         32        0.2       16384     0          0         0.0        0      
13:48:54   +ASM1    BDT10_1       DATA                                      2         32        0.2       16384     0          0         0.0        0      
13:48:54   +ASM1    BDT10_1       DATA          HOST31                      0         0         0.0       0         0          0         0.0        0      
13:48:54   +ASM1    BDT10_1       DATA          HOST31       HOST31CA0D1C   0         0         0.0       0         0          0         0.0        0      
13:48:54   +ASM1    BDT10_1       DATA          HOST31       HOST31CA0D1D   0         0         0.0       0         0          0         0.0        0      
13:48:54   +ASM1    BDT10_1       DATA          HOST32                      2         32        0.2       16384     0          0         0.0        0      
13:48:54   +ASM1    BDT10_1       DATA          HOST32       HOST32CA0D1C   0         0         0.0       0         0          0         0.0        0      
13:48:54   +ASM1    BDT10_1       DATA          HOST32       HOST32CA0D1D   2         32        0.2       16384     0          0         0.0        0      
13:48:54   +ASM1    BDT10_1       FRA                                       0         0         0.0       0         0          0         0.0        0      
13:48:54   +ASM1    BDT10_1       FRA           HOST31                      0         0         0.0       0         0          0         0.0        0      
13:48:54   +ASM1    BDT10_1       FRA           HOST31       HOST31CC8D0F   0         0         0.0       0         0          0         0.0        0      
13:48:54   +ASM1    BDT10_1       FRA           HOST32                      0         0         0.0       0         0          0         0.0        0      
13:48:54   +ASM1    BDT10_1       FRA           HOST32       HOST32CC8D0F   0         0         0.0       0         0          0         0.0        0      

and so on...

Now let’s produce the csv file with the csv_asm_metrics utility. Let’s see the help:

./csv_asm_metrics.pl -help

Usage: ./csv_asm_metrics.pl [-if] [-of] [-d] [-help]

  Parameter         Comment
  ---------         -------
  -if=              Input file name (output of asm_metrics)
  -of=              Output file name (the csv file)
  -d=               Day of the first snapshot (YYYY/MM/DD)

Example: ./csv_asm_metrics.pl -if=asm_metrics.txt -of=asm_metrics.csv -d='2014/07/04'

and generate the csv file that way:

./csv_asm_metrics.pl -if=asm_metrics.txt -of=asm_metrics.csv -d='2014/07/04'

The csv file looks like:

Snap Time,INST,DBINST,DG,FG,DSK,Reads/s,Kby Read/s,ms/Read,By/Read,Writes/s,Kby Write/s,ms/Write,By/Write
2014/07/04 13:48:54,+ASM1,BDT10_1,DATA,HOST31,HOST31CA0D1C,0,0,0.0,0,0,0,0.0,0
2014/07/04 13:48:54,+ASM1,BDT10_1,DATA,HOST31,HOST31CA0D1D,0,0,0.0,0,0,0,0.0,0
2014/07/04 13:48:54,+ASM1,BDT10_1,DATA,HOST32,HOST32CA0D1C,0,0,0.0,0,0,0,0.0,0
2014/07/04 13:48:54,+ASM1,BDT10_1,DATA,HOST32,HOST32CA0D1D,2,32,0.2,16384,0,0,0.0,0
2014/07/04 13:48:54,+ASM1,BDT10_1,FRA,HOST31,HOST31CC8D0F,0,0,0.0,0,0,0,0.0,0
2014/07/04 13:48:54,+ASM1,BDT10_1,FRA,HOST32,HOST32CC8D0F,0,0,0.0,0,0,0,0.0,0
2014/07/04 13:48:54,+ASM1,BDT10_1,REDO1,HOST31,HOST31CC0D13,0,0,0.0,0,0,0,0.0,0
2014/07/04 13:48:54,+ASM1,BDT10_1,REDO1,HOST32,HOST32CC0D13,0,0,0.0,0,0,0,0.0,0
2014/07/04 13:48:54,+ASM1,BDT10_1,REDO2,HOST31,HOST31CC0D12,0,0,0.0,0,0,0,0.0,0
2014/07/04 13:48:54,+ASM1,BDT10_1,REDO2,HOST32,HOST32CC0D12,0,0,0.0,0,0,0,0.0,0
2014/07/04 13:48:54,+ASM1,BDT11_1,DATA,HOST31,HOST31CA0D1C,0,0,0.0,0,0,0,0.0,0
2014/07/04 13:48:54,+ASM1,BDT11_1,DATA,HOST31,HOST31CA0D1D,0,0,0.0,0,2,16,0.5,8448

As you can see:

1. The day has been added (to create a date) and next ones will be calculated (should the snaps cross multiple days).
2. Only the rows that contain all the fields have been recorded into the csv file (The script does not record the other ones as they represent aggregated values).

Now I can import this csv file into Tableau.

You can imagine a lot of graphs thanks to the measures collected (Reads/s, Kby Read/s, ms/Read, By/Read, Writes/s, Kby Write/s, ms/Write, By/Write) and all those dimensions (Snap Time, INST, DBINST, DG, FG, DSK).

Let’s graph the throughput and latency per failgroup for example.

Important remark regarding some averages computation/display:

The ms/Read and By/Read measures depend on the number of reads. So the averages have to be calculated using Weighted Averages. (The same apply for ms/Write and By/Write).

Let’s create the calculated field in Tableau for those Weighted Averages:

so that weighted Average ms/Read is:

Weighted Average By/Read:

The same way you have to create:

• Weighted Average ms/Write = sum([ms/Write]*[Writes/s])/sum([Writes/s])
• Weighted Average By/Write = sum([By/Write]*[Writes/s])/sum([Writes/s])

Now let’s display the average read latency by Failgroup (using the previous calculated weighted average):

Drag the Snap Time dimension to the “columns” shelf and choose “exact date”:

Drag the Weighted Average ms/Read calculated field to the “Rows” shelf:

Drag the FG dimension to the “Color Marks” shelf:

So that the graph looks like:

Create the same graph for the “Kby Read/s” measure (except that I want to see the sum (i.e the throughput and not the average) and put those 2 graphs into the same dashboard:

Here we are.

Conclusion:

• We can create a csv file from the output of the asm_metrics utility thanks to csv_asm_metrics.

• To do so, we have to collect all the fields of asm_metrics with those options:
  • -show=inst,dbinst,fg,dg,dsk for ASM >= 11g
  • -show=inst,fg,dg,dsk for ASM < 11g

• Once you uploaded the csv file into your favourite visualization tool, don’t forget to calculate weighted averages for ms/Read, By/Read, ms/Write and By/Write if you plan to graph the averages.

• You can imagine a lot of graphs thanks to the measures collected (Reads/s, Kby Read/s, ms/Read, By/Read, Writes/s, Kby Write/s, ms/Write, By/Write) and all those dimensions (Snap Time, INST, DBINST, DG, FG, DSK).

You can download the csv_asm_metrics utility from this repository or copy the source code from this page.

UPDATE: You can see some use cases here.

Visualize the IO source thanks to Tableau and AWR

As you know, the wait event “db file sequential read” records “single block” IO performed outside the database buffer cache. But does the IO come from:

• Filesystem cache (If any and used)
• Disk Array cache
• SSD
• Spindle Disks
• …..

It could be interesting to visualize the distribution of the IO source:

• Should you migrate from a cached filesystem to ASM (You may need to increase the database cache to put the previous Filesystem cached IOs into the database cache).
• Should you use Dynamic Tiering and want to figure out where the IOs come from (SSD, Spindle Disks..).

To do so, I’ll use the AWR data coming from the dba_hist_event_histogram view and Tableau. I’ll also extract the data coming from dba_hist_snapshot (to get the begin_interval_date time).

alter session set nls_date_format='YYYY/MM/DD HH24:MI:SS';
alter session set nls_timestamp_format='YYYY/MM/DD HH24:MI:SS';

select * from dba_hist_event_histogram where
snap_id >= (select min(snap_id) from dba_hist_snapshot
where begin_interval_time >= to_date ('2014/06/01 00:00','YYYY/MM/DD HH24:MI'))
and event_name='db file sequential read';

select * from dba_hist_snapshot where begin_interval_time >= to_date ('2014/06/01 00:00','YYYY/MM/DD HH24:MI');

As you can see, there is no computation. This is just a simple extraction of the data.

Then I put those data into 2 csv files (awr_snap_for_june.csv and awr_event_histogram.csv).

1) Now, launch Tableau and select the csv files and add an inner join between those files:

2) Go to the worksheet and put the “begin interval time” dimension into the “column” and change it to an “exact date” (Instead of Year):

3) Put the “Wait count” measure into the “Rows” and create a table calculation on it:

Choose “difference” as the “WAIT_COUNT” field is cumulative and we want to see the delta between the AWR’s snapshots.

4) My graph now looks like:

The Jun 14 and Jun 20 the database has been re-started and then the difference is < 0.

5) Let’s modify the formula to take care of database restart into the delta computation:

Customize

Name: “Delta Wait Count” and change ZN(SUM([Wait Count])) – LOOKUP(ZN(SUM([Wait Count])), -1) to max(ZN(SUM([Wait Count])) – LOOKUP(ZN(SUM([Wait Count])), -1),0):

So that now the graph looks like:

6) Now we have to “split” those wait count into 2 categories based on the wait_time_milli measure coming from dba_hist_event_histogram. Let’s say that:

• “db file sequential read” <= 4 ms are not coming from spindle disks (So from caching, SSD..).
• “db file sequential read” > 4 ms are coming from spindle disks.

Let’s implement this in tableau with a calculated field:

Name: “IO Source” and use this formula:

Feel free to modify this formula according to your environment.

Now take the “IO Source” Dimension and put it into the Color marks:

So that we can now visualize the IO source repartition:

 

Remarks:

• Karl Arao presented another example usage of Tableau into this blog post.
• Should you need to retrieve “db file sequential read” buckets < 1 ms, then you can use oracle_trace_parsing from Kyle Hailey.

Update 1: Example of oracle_trace_parsing usage into “Oracle “Physical I/O” ? not always physical” blog post.

Update 2: Another way to retrieve “db file sequential read” buckets < 1 ms (With external tables this time) into Nikolay Savvinov blog post.