Who's online

There are currently 0 users and 28 guests online.

Recent comments


Pimp my collectl-advanced system monitoring using collect-utils part I

I have recently written about collectl, a truly superb troubleshooting utility, in a previous post. After comments from Mark Seeger (the author) and Kevin Closson (who has used it extensively and really loves it), I have decided to elaborate a bit more about what you can do with collectl.

Even though it’s hard to believe, collectl’s functionality can be extended by using the collectl-utilities from sourceforge, available here:

Like collectl, you can either download a source tgz file or a noarch-RPM. Collectl-utils consist of three major tools, out of which I’d like to introduce the first one: colplot. When finding time I’ll create a post about the other part, most likely about colmux first.


I mentioned in said previous post that you can use the “-P” option to generate output in a plot format. This in turn can be fed to your favourite spreadsheet application, or alternatively into gnuplot. When chosing to use a spreadsheet application, it’s your responsibility to decide what to do with the raw data, each time you load a plotfile. Maybe, one day I’ll write a collectl-analyzer which does similar things to nmon-analyzer, but that has to wait for now. So if you are lazy like me, you need another alternative, and it comes easily accessible in the form of gnuplot.

Although I am very impressed by what gnuplot can do, I never had the time or energy to get to grips with all its options. When at University I used Mathematica 2 (yes it’s been a while) and thought the plot2d() function was complex …

Now for the good news: the complexity of gnuplot is nicely hidden by colplot, which takes the burden of generating the plot files away from the user. And to make it more comfortable, all of this happens through a web interface. All you need is a web server such as apache and a little bit of initial configuration for it to work. I should also note that colplot can be used on the command line as well, but that is out of scope of this article.

This time around I downloaded the source tarball rather than the RPM as I wanted more control over the installation process. If you chose the RPM it is good to know that it has intelligence to tell SLES apart from RHEL and updates the web server configuration accordingly. If you decide to manually install colplot, check the INSTALL script as it can help you getting started. And don’t forget to read INSTALL-colplot and consult colplot-apache.conf for a sample apache configuration. The latter can go to /etc/httpd/conf.d on RHEL and will take effect after a reloading of the apache configuration. You also need collectl installed on the host running the collect-utils.

Colplot uses a directory, usually called plotfiles, where the recorded collectl output is stored. By default, it resides in the same directory as colplot but can be changed in the GUI.

I am thinking of using NFS to export the plotfiles directory, so that each monitored host could mount the directory and store output files. The more progressive use of SSHFS is probably out of scope for most database servers, but on my lab I’m king and do what I like. I personally found it easiest to use “collectl -P -f /mnt/sshfs/plotfiles/ ” to generate the files, where mount/sshfs/plotfiles was mounted from the web server host. If you are planning on generating the colplot output file names manually, i.e. not pointing to a directory, then make sure they are unique! This makes is easy to compare systems, as we’ll see below. One thing I noticed is that detail files all get their own trace file name in the form “host-date.type”, where type is dsk for detailed disk information etc.

After all the webserver setup is complete, point your browser to the host where you installed colplot. As I said, the “plotfiles” directory is scanned for files, which are processed. You see the following screen:

Using it

In the GUI, the first step you define which time of day with matching/gathered collectl information you would like to visualise (open the above screenshot in a separate window to better follow this discussion).

  • You can limit the information to be displayed to a certain time period, i.e. if you captured a day’s worth of statistics but only need the hour from 13:00 to 14:00 that’s a simple setting in the user interface
  • Alternatively, select “last 60 minutes” for the most recent period

You can also list the contents of the plotfiles directory, or even change the location-but bear in mind that the webserver still has to be able to read files from there!

If you like, you can instruct colplot to narrow down the files to be plotted by editing “filenames containing”. If the plotfiles directory contains information to satisfy the period/names you are interested in, it will plot it after a click on “Generate plot”. I suggest a display of “SysPlot” initially, which plots the systems recorded in the colplot files side by side. This is very useful for comparison of system health, especially in clusters. You should experiment with the different plot settings, which are very useful to do all sorts of analysis and allow aggregation on days, system and plots and various combinations of these. By the way the system name is derviced from the hostname when using the colplot -P -f /path/to/plotfiles/ … command.

Once you familiarised yourself with the options, you can further narrow down which data you interested in. I would suggest the “All Plots” option to get you started, unless of course you know what you are after. Colplot, like collectl differentiates between “summary” and “detail” plots. Of course, it can only plot what you recorded! Each of these has a convenient “All Plots” option to display all the information gathered. Here’s the slightly cropped output from a 2 node cluster I gathered (click for a larger view):

A very usful function is to email the results either in PDF or PNG format. For this to work you need uuencode (package sharutils in RHEL) on the web server host, and you need to be able to send email via the command line-colplot uses the mail (1) utility to send email. Sending a PDF is probably more useful than the PNG option, as the latter will send a each graph separately in a tar archive.


I can’t say how impressed I am with colplot, it’s really great for working out what happened when during a benchmark. The great things is the comparison of systems side by side which gives clear indications of imbalances and trensing. Using colplot is also a lot easier than writing your own spreadsheet macros to visualise the data. I really like it!

Full scans, direct path reads and ORA-8103 error hacking session video here (plus iTunes podcast address!)

I have uploaded the latest hacking session video to I have edited it a little, I cut out the part where I spilled an entire Red Bull onto my desk, with some onto my laptop (some keys are still sticky:)

Also, I do upload all these sessins into iTunes – so you can subscribe to my podcast! That way you can download the videos into your computer, phone or iPad. I have deliberately used 1024×768 resolution so it would look awesome on iPad screen! (so hopefully your commute time gets a bit more fun now ;-)



iTunes video-podcast:



Pickler Fetch – What is it?

How many of you have seen the step COLLECTION ITERATOR PICKLER FETCH appearing in Explain Plans and wondered “what is a Pickler Fetch”?

{Note – make sure you check out the comments on this post if you are interested in Pickler Fetches. Timur Akhmadeev and Dom Brooks correct and extend my original efforts}.

| Id  | Operation                             | Name       | Rows  | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT                      |            |       |       |    31 (100)|          |
|*  1 |  FILTER                               |            |       |       |            |          |
|   2 |   NESTED LOOPS                        |            |       |       |            |          |
|   3 |    NESTED LOOPS                       |            |     1 |    71 |    31   (4)| 00:00:01 |
|   4 |     SORT UNIQUE                       |            |     4 |     8 |    26   (0)| 00:00:01 |
|   5 |      COLLECTION ITERATOR PICKLER FETCH|            |     4 |     8 |    26   (0)| 00:00:01 |
|*  6 |     INDEX RANGE SCAN                  | CHHE_PK    |     1 |       |     2   (0)| 00:00:01 |
|   7 |    TABLE ACCESS BY INDEX ROWID        | CHILD_HEAP |     1 |    69 |     4   (0)| 00:00:01 |

It is where Oracle works through a SQL array. I’ve usually seen this within PL/SQL where an array variable is used within a procedure or package, being passed in containing a set of accounts or customers that are of interest and is used with a cursor. But you might also see it as the step processing the output of a pipelined function (one that returns a set of “rows”) being cast into a table. See this example by Tom Kyte. Tom also suggests that it is called a “Pickler” fetch as the data is pickled – packed and formatted. I never knew that, I just thought someone in Oracle development was having a giggle and it was “Pickled” as it was preserved from the PL/SQL side of the SQL engine. It seems that I was a little off-target with that.

{My thanks to Timur (see comments) who corrected me when I said it was a PL/SQL array. It is not, it is a SQL object}.

The above plan was generated for the below SQL statement:

cursor c_collect is
select  PARE_ID
from child_heap
where pare_id in (select  id  from table(p_parents) p)
and   cre_date >=trunc(p_date) and cre_date < trunc(p_date+1);

Anyway, I just wanted to make a few comments about Pickler Fetches:

  • The CBO is not very good at looking “inside” such arrays to determine the cardinality of that step. This can cause performance issues.
  • Because using them involves a few steps and potentially involves grants, it is possible for them to be “temporarily removed” during testing and so their impact is not seen.
  • They can leak memory. I think.

I’m now going to cover each of those points in turn. If you just wanted to know what a pickler fetch is and I’ve answered that for you, I suggest you go back to whatever you were doing before you started reading this :-)

By “not very good at looking inside pickler fetches” I mean that the CBO seems to make a poor “base it on the first seen array” (V11 up) or utterly fails to identify how many records are inside the SQL array (V10 down), depending on the Oracle version. From an Oracle system I’ve got the two following examples:

| Id  | Operation                               | Name              | Rows| Bytes | Cost (%CPU)| Time     | Pstart
|   9 |      NESTED LOOPS                       |                   |   1 |   139 |   114   (1)| 00:00:01 |      
|* 10 |       HASH JOIN SEMI                    |                   |   1 |    50 |    11  (10)| 00:00:01 |      
|* 11 |        TABLE ACCESS BY INDEX ROWID      | ABCDE_ABCDEFGHIJ  |   3 |   144 |     4   (0)| 00:00:01 |      
|* 12 |         INDEX RANGE SCAN                | XX_PK             |   3 |       |     3   (0)| 00:00:01 |      
|  13 |        COLLECTION ITERATOR PICKLER FETCH|                   |     |       |            |          |      
|  14 |       PARTITION RANGE ITERATOR          |                   |   6 |   534 |   103   (0)| 00:00:01 |   KEY
|* 15 |        INDEX RANGE SCAN                 | AAXX_PK           |   6 |   534 |   103   (0)| 00:00:01 |   KEY

Note the cost and expected rows for the Pickler Fetch step. Or rather, the lack of them.

This would manifest itself in the following way in OEM screens:

Note the cardinality of the Pickler Fetch step is 0 {sorry, I cut off the column headings}. This resulted in this step having a cost of 0 and all the subsequent steps of having an expected cardinality of one and having very low costs – between 1 and 3 {Again, not shown, sorry}.

The end result of this was that the CBO struggled to accuratley cost any given plan and came up with several, usually quite terrible, plans that it swapped between as other table stats varied. The CBO was picking between very complex plans with total costs of 100 or 200 or so! Any difference was “significant”.

Please note, OPTIMIZER_DYNAMIC_SAMPLING was set to 4 on this system and I tried hints and session settings to higher levels and they did not prompt the CBO to look into the array, on at least.

In 11.1 things seem to be better, as is shown in the explain plan at the top of this post. The step has a cost. I have to confess, I have not tested this very much on 11 {and if anyone has, feel free to correct me/add enlightenment via comments or pointers to other sources}, but it seems to set the cardinality to the number of elements the Pickler Fetch finds in the first itteration. Unless it uses the same sort of trick Oracle 11 now uses for bind variables (detecting when the value supplied is out of range and generates a new plan) this is going to lead to the old and much ‘loved’ issue of the plan being fixed by the first execution, irrespective of how suitable that plan is.

How do you fix this issue? Well, I resort to the cardinality hint. Usually the number of records being passed into the array is not too variable and any half-decent value is better than nothing in Oracle 10 and before. As for in 11, I like stating the value rather than risking a variable ‘first seen at parsing time’ setting. It is a judgement call. The below is from 11.1 but I’ve used it extensively in 10.2, where the impact is much more significant:

select  PARE_ID
from child_heap
where pare_id in (select  id  from table(p_parents) )
and   cre_date >=trunc(p_date) and cre_date < trunc(p_date+1)
| Id  | Operation                             | Name       | Rows  | Bytes |
|   0 | SELECT STATEMENT                      |            |       |       |
|*  1 |  FILTER                               |            |       |       |
|   2 |   NESTED LOOPS                        |            |       |       |
|   3 |    NESTED LOOPS                       |            |     1 |    71 |
|   4 |     SORT UNIQUE                       |            |     4 |     8 |
|   5 |      COLLECTION ITERATOR PICKLER FETCH|            |     4 |     8 |
|*  6 |     INDEX RANGE SCAN                  | CHHE_PK    |     1 |       |
|   7 |    TABLE ACCESS BY INDEX ROWID        | CHILD_HEAP |     1 |    69 |
-- and now with hint
select  PARE_ID
from child_heap
where pare_id in (select /*+ CARDINALITY(p,11) */ id  from table(p_parents) p)
and   cre_date >=trunc(p_date) and cre_date < trunc(p_date+1)
| Id  | Operation                             | Name       | Rows  | Bytes |
|   0 | SELECT STATEMENT                      |            |       |       |
|*  1 |  FILTER                               |            |       |       |
|   2 |   NESTED LOOPS                        |            |       |       |
|   3 |    NESTED LOOPS                       |            |     1 |    71 |
|   4 |     SORT UNIQUE                       |            |    11 |    22 |
|   5 |      COLLECTION ITERATOR PICKLER FETCH|            |    11 |    22 |
|*  6 |     INDEX RANGE SCAN                  | CHHE_PK    |     1 |       |
|   7 |    TABLE ACCESS BY INDEX ROWID        | CHILD_HEAP |     1 |    69 |

Note the change of ROWS to 11 in step 5. In V10 this is a change from blank to 11 and in real situations, do not be at all suprised if the plan changes dramatically – away from nested loop access and more to hash joins. {I should note, the cardinality hint is not documented in Oracle 10 or 11 and any use you make of it in live code is your responsibility. Sorry about that}.

What about my second point, about testing them? Well, as an example of testing Pickler processing of SQL arrays, which are defined SQL types, this is what I had to do to run my basic test:

-- pare_types_demo.sql
-- create and use some record and table types.
create or replace type pare_rec as object
(id number(10)
,name varchar2(100))
create or replace type pare_rec_t
as table of pare_rec
set serveroutput on size unli
create or replace procedure get_children
(p_parents in pare_rec_t
,p_date    in date)
cursor c_collect is
select  PARE_ID
from child_heap
where pare_id in (select id from table(p_parents))
and   cre_date >=trunc(p_date) and cre_date < trunc(p_date+1);
  for v_rec in c_collect
    dbms_output.put_line('parent '||to_char(v_rec.pare_id)||' on '||to_char(v_rec.cre_date)
    ||' value '||v_rec.vc_1);
  end loop;
l_date        date           :=sysdate-100;
l_parents     pare_rec_t;
  l_parents :=pare_rec_T(PARE_REC(1,'ERIC')

I had to create a table type, which is the SQL array, and this was based on an object type which I had to create first {you can have table types based on standard SQL types but very often they are based on a “row” object}. After creating the stored procedure, I had to define and populate the array with a set of records which I then passed in to my procedure call . {If you want to repeat this yourself, check out my postings on IOTs to get the table creation statement for table CHILD_HEAP}.
Now, I created those types so I had access to them. If those types do not belong to you you have to be granted execute on the types to reference them. Not select, execute. Some sites have a pretty strict attuitude to granting execute on anything and types seem to get forgotten when the execute priviledges against packages and procedures are set up. In a recent situation I had, I was forced to do some testing work on Live and it had taken people with big sticks to get me select access on data. Execute privileges were totally refused. Calmly explaining why it was needed and how it was acceptable fell on not so much deaf as bricked-up ears.

So, for testing, the reference to an array passed in is often replaced by a little sub-select. After all, quite often what is being passed in for a pickler fetch is actually a set of records {but a subset of the rows} from a table that has been collected by a previous processing step.
As an example of such a change:

where  ev.SYS_ID                       = pag.SYS_ID
and    ev.EXT_ID                       = pag.EXT_ID
and    P_EXT_ID                        = pag.EXT_ID
and    P_SYS_ID                        = pag.SYS_ID
and    ev.CREATED_DT                  >= pag.START_DT
and (ev.ID, ev.TX) in (select ID,TX from table(P_ARRAY_IN))
where  ev.SYS_ID                       = pag.SYS_ID
and    ev.EXT_ID                       = pag.EXT_ID
and    P_EXT_ID                        = pag.EXT_ID
and    P_SYS_ID                        = pag.SYS_ID
and    ev.CREATED_DT                  >= pag.START_DT
--and (ev.ID, ev.TX) in (select ID,TX from table(P_ARRAY_IN)) --NB swap back in before release!

All that has changed is that we now have a little sub-select rather than the casting of the SQL array into a table and, heck, as the developer might say, those were the records that would have been passed in, the code still works as expected and the same data comes back. No identifying which object types you need, no getting the execute permissions, no populating it yourself in the test harness, we can just swap back in the array later.

{I apologise to good developers, who even now are throwing imaginary darts at me. I know Dawn, you would certainly not do this. But I’ve seen it a couple of times. Developers have enough on their plate to go worrying about esoteric aspects of the CBO}

But the thing is, Oracle can look at that select and evaluate it’s cost and get an expected cardinality. The pickler fetch version has the issues I’ve just covered. I had to deal with a situation just like this last year, it does happen. In dev it was fine, in pre-live testing it was not.

What about memory leaks? Well, I had lots of issues with SQL arrays and memory leaks with oracle 10.1. and 10.2 at one client site and there is a documented bug in oracle 8 with pickler fetch and memory leaks but I have to confess, a quick metalink search did not find any hits for Oracle 10 and 11. So maybe you should not trust me on that one. In the situation I saw the arrays were massive, several MB at times, and so if you are using SQL arrays to pass in a modest list of eg accounts or customers, it is not going to be an issue anyway.

You know, this was just going to be a quick post on something I’ve been meaning to mention for months, not a small essay :-) .

Secret hacking session – full scans, direct path reads, object level checkpoints, ORA-8103s! (again)

I’m mentioning this again just in case you missed the announcement (because I posted it on the weekend):


There will be anotner free Secret hacking session – about full scans, direct path reads, object level checkpoints, ORA-8103s!

It will happen tomorrow, Tuesday 9th August, online!


Register here:

See you soon!


Are you getting the most out of your Exadata performance? Part 1

In almost all of the Exadata migration projects I’ve been part of, the client sees immediate speedup & performance increase when testing their workload on Exadata (of course, we’ve made sure that we do plan & execute the tasks right). However, my performance geek’s nature usually doesn’t allow to stop there and leave the client with just 2x or 3x performance increase. For data warehousing and reporting workloads, Exadata can do much better than just 2-3x performance increase! 

This is why I will write this article series about Getting the Most out of your Exadata Performance. I will write a bunch of random articles, based on my experience and lessons learned – and some day I may consolidate it all into a more formal paper.

So, here’s the first article (PDF format).

IOT Part 4 – Greatly Boosting Buffer Cache Efficiency

<..IOT1 – the basics
<….IOT2 – Examples and proofs
<……IOT3 – Significantly reducing IO
IOT5 – Primary Key issues..>

So far I have covered the basics of Index Organized Tables, created some examples and shown how IOTs can help reduce the cost of single row selects and then how they can greatly aid range scans. Follow the above links for details.

Now I’d like to show that the use of IOTs has the potential to make your block buffer cache (BBC) far more efficient. Going to disc is very,very slow compared to going to memory {NB solid state storage improves this situation but does not remove it}. The block buffer cache has always been critical to oracle SQL Select performance as it allows you to access data in memory rather than disc and in general the more block buffer cache you have the faster your system will be.
{I am of the opinion that the BBC is even more important now than ever. As hard discs get larger we are seeing fewer and fewer spindles per GB of storage and, in essence, disc storage is effectively getting slower – because more data is hosted on the same number of spindles and those spindles are not themselves getting faster – I digress, for more details see posts Big Discs are Bad and IOPs and Form Factors}

In the scenario I’ve covered in my previous posts on IOTs we have a system where child data is coming in for many parents every day for 100 days. With a heap table the data pours into the growing end of the table, usually a record or two per parent each day and no guarantee that if two records come in, they will be put into the same block.

So, when you select a child record for a parent you get the situation shown below:

For many systems, the Block Buffer Cache is holding a lot of data no queries asked for -collateral data

When oracle needs to collect a record from the table, it has to read the whole block. Oracle only reads in tablespace data in whole blocks. That record comes with many other records in it that you did not ask for or want. I refer to this as Collateral Data – innocent bystander data that has got pulled into the BBC just because it was in the same block as required data. The larger the block size, the more collateral data there is.

To get all the child records for the parent, you need to read all those individual table blocks holding one or two records of interest. For our 100 child records you will probably need to read in close to 100 table blocks. Your Block Buffer Cache is filling up with of blocks where only one row out of each block is “of interest”. If that is one row out of 80 in a block, you are effectively wasting 98.75% of the space that table takes up in the block buffer cache.

With an IOT the situation is very different. We have already seen in my previous post on reducing IO that for a range scan on the IOT, oracle does not need to go and collect records from blocks scattered throughout the table. It simply collects the IOT leaf blocks holding the relevant data. Not only does this require less IO, it also results in the fetched blocks mostly holding the required data. The percentage of collateral data is greatly reduced:

IOTs are a powerful tool in reducing collateral data and using the BBC more efficiently

Thus instead of 100 table blocks that mostly hold collateral data, you have 2 or 3 blocks holding mostly the data you are interested in. Your wastage, the collateral data, is about 33-50%. With my example tables from post IOT2, it is indeed an average of 2-3 IOT blocks holding all the data for a single parent and 100 heap table blocks holding the same data.

You can think of it another way.

With my IOT I use only 3% of the memory to cache a single set of records for a parent compared to that needed with a normal HEAP table.

Let’s extend that idea a little. Let’s say I have 100,000 customer and 5% of the customers are active.
Each customer has on average 500 * 200-byte activity records for 100K of data
Each 8K Heap Table block holds 40 records, a very inefficient IOT holds 20 records per block {I’ll go into the details of this in a later post, but I am being a little unfair on the IOT here}.
There is some grouping of records for the customer so that those 500 records are over 400 blocks.
The whole working set of 400 blocks * (100,000*0.05) customers *8K = 2,000,000K

Now replace the table with an IOT keyed on customer ID.
Those 400 records would be spread over (400/20)+2 blocks. The +2 is the start and end of the range.
That is 22 *(100,000*0.05) *8K = 110,000K

Both would also need the overhead of an index structure to be cached also, for the Heap table it is the Primary Key index, for the IOT it is the rest of the IOT structure.

So you would need 2GB or so of Block Buffer Cache dedicated to caching the working set held as a heap table and 110MB of Block Buffer Cache dedicated to caching the IOT equivalent.

How many of you spotted that the space needed to “cache the heap table working set” is actually twice the size of the table? Well, that is because with a 5% working set and 40 records per block, there is a fair chance that some of those Collateral Data records in each block are for the 5% working set. Over all, almost every block will hold data for two or three active users. This is one of the complications of working out how efficient you block buffer cache is likely to be and I’ll have to leave that to another post. In reality you would need to hold 99% of the heap table in memory to cache the whole working set, so still 1GB. The IOT is still far more efficient.

I’ll just finish by saying that on one project I worked on we would have needed several hundred GB of Block Buffer Cache to hold the working set of the main tables if held as heap tables. That volume of memory was simply not available. With IOTs this reduced to about 40GB. This was available. The majority of this working set was able to stay in their SGA and it meant that so long as the instance stayed up, that working set of data mostly stayed in cache. I was able to see on the live system that processing of the data for these active customers was mostly being supported by consistent gets and less than 1% disk reads.

So, by using IOTS I reduced consistent gets dramatically, made more efficient use of the block buffer cache and, as a result of that, reduced the number of physical reads needed to support the consistent gets.

An introduction to collectl

Some of you may have seen on twitter that I was working on understanding collectl. So why did I start with this? First of all, I was after a tool that records a lot of information on a Linux box. It can also play information back, but this is out of scope of this introduction.

In the past I have used nmon to do similar things, and still love it for what it does. Especially in conjunction with the nmon-analyzer, an Excel plug in it can create very impressive reports. How does collectl compare?

Getting collectl

Getting collectl is quite easy-get it from sourceforge:

The project website including very good documentation is available from sourceforge as well, but uses a slightly different URL:

I suggest you get the archive-independent RPM and install it on your system. This is all you need to get started! The impatient could type “collectl” at the command prompt now to get some information. Let’s have a look at the output:

$ collectl
waiting for 1 second sample...
#cpu sys inter  ctxsw KBRead  Reads KBWrit Writes   KBIn  PktIn  KBOut  PktOut
1   0  1163  10496    113     14     18      4      8     55      5      19
0   0  1046  10544      0      0      2      3    164    195     30      60
0   0  1279  10603    144      9    746    148     20     67     11      19
3   0  1168  10615    144      9    414     69     14     69      5      20
1   0  1121  10416    362     28    225     19     11     71      8      35

The “ouch” has been caused by my CTRL-c to stop the execution.

Collectl is organised to work by subsystems, the standard option is to print CPU, disk and network subsystem, aggregated.

If you don’t know what information you are after, you could use the –all flag to display aggregated information across all subsystems. Be warned that you need a large screen for all that output! For even more output, add the –verbose flag to the –all option and you need a 22” screen at least. The verbose flag prints more output, as the name suggests. For the disk subsystem you can view the difference:

$ collectl -sd -i 5 --verbose
waiting for 5 second sample...

#KBRead RMerged  Reads SizeKB  KBWrite WMerged Writes SizeKB
162     136     10     15      187      30     19      9
109      24      9     11      566     118     23     24
$ collectl -sd -i 5
waiting for 5 second sample...
#KBRead  Reads KBWrit Writes
9865     73    190     23

Each subsystem can be queried individually, the default monitoring interval is 1 second. The man page for collectl lists the following subsystems:


b - buddy info (memory fragmentation)
c - CPU
d - Disk
f - NFS V3 Data
i - Inode and File System
j - Interrupts
l - Lustre
m - Memory
n - Networks
s - Sockets
t - TCP
x - Interconnect
y - Slabs (system object caches)

As the name suggests, these sub systems provide summary information. Summaries are ok for a first overview, but don’t forget that information is aggregated and detail is lost.

From an Oracle point of view I’d probably be most interested in the CPU, disk and memory usage. If you are using RAC, network usage can also be interesting.

Detailed subsystem information is available for these (again taken from the excellen manual page):

D - Disk
E - Environmental data (fan, power, temp),  via ipmitool
F - NFS Data
J - Interrupts
L - Lustre OST detail OR client Filesystem detail
N - Networks
T - 65 TCP counters only available in plot format
X - Interconnect
Y - Slabs (system object caches)
Z - Processes

You can combine subsystems, and you can combine detail and summary information. Bear in mind though that this becomes a lot of information for a putty session of gnome-terminal!

In interactive mode, you might want to consider the –home flag, which does a top-like refresh and prints real time information without scrolling: very neat!

But even with the –-home option, digesting all that information visually can be a bit daunting, which leads me to my next section.

Generating graphical output

While all the textual information is all nice and good, it is difficult to visualise. Collectl can help you with that as well. All you need to do is generate a file in tab format, which is as simple as adding the –P and –f options. Since you can’t be overwhelmed with the information gathered in a file (unlike on standard out), you could use the detail switches. If you have the luxury, create the file with the information in a directory expored via samba and analyse it with Excel or other utilities. It’s possible to use gnuplot as well, but I found that a bit lacking for interactive use. The collectl-utils provide a CGI script to analyse collectl files on the host which can be convenient. Here is an example for measuring CPU, memory and all disks with a monitoring interval of 15 seconds. The file will be in “Plot” format (-P) and goes to /export/collectl/plotfiles:

$ collectl -sdmn -i 15 -P -f /export/collectl/plotfiles

Note that you can’t use the verbose flag here, and you also shouldn’t use a file name with the –f switch!

The resulting file is called After renaming it to hostname-yyyymmdd.txt it can quite easily be imported using your favourite spreadsheet application. Imagine all the graphs you could produce with it! Also the header contains interesting information:

# Collectl:   V3.5.1-1  HiRes: 1  Options: -sdmn -i 15 -P -f /export/collectl/plotfiles
# Host:       node1 DaemonOpts:
# Distro:     Red Hat Enterprise Linux Server release 5.5 (Tikanga)  Platform:
# Date:       20110805-142647  Secs: 1312550807 TZ: +0100
# SubSys:     dmn Options: z Interval: 1 NumCPUs: 16  NumBud: 0 Flags: i
# Filters:    NfsFilt:  EnvFilt:
# HZ:         100  Arch: x86_64-linux-thread-multi PageSize: 4096
# Cpu:        AuthenticAMD Speed(MHz): 2210.190 Cores: 4  Siblings: 4
# Kernel:     2.6.18-194.el5  Memory: 65990460 kB  Swap: 16809976 kB
# NumDisks:   173 DiskNames: c0d0 sda sdb sdc sdd sde sdf sdg sdh sdi sdj sdk sdl sdm sdn sdo sdp sdq sdr sds sdt sdu sdv sdw sdx sdy sdz sdaa sdab sdac sdad sdae sdaf sdag sdah sdai sdaj sdak sdal sdam sdan sdao sdap sdaq sdar sdas sdat sdau sdav sdaw sdax sday sdaz sdba sdbb sdbc sdbd sdbe sdbf sdbg sdbh sdbi sdbj sdbk sdbl sdbm sdbn sdbo sdbp sdbq sdbr sdbs sdbt sdbu sdbv sdbw sdbx sdby sdbz sdca sdcb sdcc sdcd sdce sdcf sdcg sdch sdci sdcj sdck sdcl sdcm sdcn sdco sdcp sdcq sdcr sdcs sdct sdcu sdcv sdcw sdcx sdcy sdcz sdda sddb sddc sddd sdde sddf sddg dm-0 dm-1 dm-2 dm-3 dm-4 dm-5 dm-6 dm-7 dm-8 dm-9 dm-10 dm-11 dm-12 dm-13 dm-14 dm-15 dm-16 dm-17 dm-18 dm-19 dm-20 dm-21 dm-22 dm-23 dm-24 dm-25 dm-26 dm-27 dm-28 dm-29 dm-30 dm-31 dm-32 dm-33 dm-34 dm-35 dm-36 dm-37 dm-38 dm-39 dm-40 dm-41 dm-42 dm-43 dm-44 dm-45 dm-46 dm-47 dm-48 dm-49 dm-50 dm-51 dm-52 dm-53 dm-54 dm-55 dm-56 dm-57 dm-58 dm-59 dm-60
# NumNets:    8 NetNames: lo: eth0: eth1: eth2: eth3: sit0: bond0: bond1:
# SCSI:       DA:0:00:00: ... DA:2:00:00:00

This should be enough to remind you of where you were running this test.

Run duration and interval

Use the –i flag to change the monitoring interval, this is the same as you’d do with SAR or iostat/vmstat and the like. You could then either use the –c option to count n samples, or alternatively use –R to run for n weeks, days, hours, minutes or seconds, each of which are abridged with their first letter. For example to run for 15 minutes with samples taken every 15 seconds, you’d say collectl –i 15 –R 15m.

Quick and dirty

If you need an interactive overview of what’s going on top-style, you can use the –top flag. This will print output very similar to the top command, but this time you have a lot more options to sort on. Use collectl –showtopops. This is so cool that I couldn’t help just listing the options here:

$ collectl --showtopopts
The following is a list of --top's sort types which apply to either
process or slab data.  In some cases you may be allowed to sort
by a field that is not part of the display if you so desire


vsz    virtual memory
rss    resident (physical) memory

syst   system time
usrt   user time
time   total time

rkb    KB read
wkb    KB written
iokb   total I/O KB

rkbc   KB read from pagecache
wkbc   KB written to pagecache
iokbc  total pagecacge I/O
ioall  total I/O KB (iokb+iokbc)

rsys   read system calls
wsys   write system calls
iosys  total system calls

iocncl Cancelled write bytes

Page Faults
majf   major page faults
minf   minor page faults
flt    total page faults

Miscellaneous (best when used with --procfilt)
cpu    cpu number
pid    process pid
thread total process threads (not counting main)


numobj    total number of slab objects
actobj    active slab objects
objsize   sizes of slab objects
numslab   number of slabs
objslab   number of objects in a slab
totsize   total memory sizes taken by slabs
totchg    change in memory sizes
totpct    percent change in memory sizes
name      slab names

Filtering information

Let’s say you are running multiple ASM disk groups in your system, but you are only interested in the performance of disk group DATA. The –sD flag will print all the information for all disks (LUNs) of the system. Collectl reports disks as the native devices and dm- devices. For multipathed devices you obviously want to look at the dm- device. You could use the multipath –ll command to map dm- device to WWIDs and your disks in the end. Let’s say you found out that the disks you need to look at are /dev/dm-{1,3,5,8} you could use the –dskfilt flag, which takes a perl regex. In my example, I could use the following command to check on those disks:

collectl -sD -c 1 --dskfilt "dm-(1\b|3\b|5\b|8\b)"
waiting for 1 second sample...

#          <---------reads---------><---------writes---------><--------averages--------> Pct
#Name       KBytes Merged  IOs Size  KBytes Merged  IOs Size  RWSize  QLen  Wait SvcTim Util
dm-1             0      0    0    0       0      0    0    0       0     0     0      0    0
dm-3             0      0    0    0       0      0    0    0       0     0     0      0    0
dm-5             0      0    0    0       0      0    1    1       0     0     0      0    0
dm-8             0      0    0    0       0      0    0    0       0     0     0      0    0

Note the “\b” boundary, which is my uneducated way to saying that the expression should match dm-1, but not 10, or anything else that extends beyond number one.

Additional filters you can apply can be found in the output of collectl –showsubopts as well as in section subsystem options in the manpage.


Used correctly, collectl is the swiss army knife for system monitoring, the level of detail which can be gathered is breathtaking. Thanks Mark Seger! And aplogies for all the good stuff I’ve been missing!

Friday Philosophy – Oracle Performance Silver Bullet

Silver Cartridge and Bullet

For as long as I have been working with Oracle technology {which is now getting towards 2 decades and isn’t that pause for thought} there has been a constant search for Performance Silver Bullets – some trick or change or special init.ora parameter {alter system set go_faster_flag=’Y'} you can set to give you a guaranteed boost in performance. For all that time there has been only one.

There are a few performance Bronze Bullets…maybe Copper Bullets. The problem is, though, that the Oracle database is a complex piece of software and what is good for one situation is terrible for another. Often this is not even a case of “good 90% of the time, indifferent 9% of the time and tragic 1% of the time”. Usually it is more like 50%:30%:20%.

Cartridge with copper bullet &spent round

I’ve just been unfair to Oracle software actually, a lot of the problem is not with the complexity of Oracle, it is with the complexity of what you are doing with Oracle. There are the two extremes of OnLine Transaction Processing (lots of short running, concurrent, simple transactions you want to run very quickly by many users) and Data Warehouse where you want to process a vast amount of data by only a small number of users. You may well want to set certain initialisation parameters to favour quick response time (OLTP) or fastest processing time to completion (DW). Favouring one usually means a negative impact on the other. Many systems have both requirements in one… In between that there are the dozens and dozens of special cases and extremes that I have seen and I am just one guy. People get their database applications to do some weird stuff.

Partitioning is a bronze bullet. For many systems, partitioning the biggest tables makes them easier to manage, allows some queries to run faster and aids parallel activity. But sometimes (more often than you might think) Partitioning can drop rather than increase query or DML performance. In earlier versions of Oracle setting optimizer_index_caching and optimizer_index_cost_adj was often beneficial and in Oracle 9/8/7 setting db_file_multiblock_read_count “higher” was good for DWs….Go back to Oracle 7 and doing stuff to increase the buffer cache hit ratio towards 98% was generally good {and I will not respond to any comments citing Connors magnificent “choose your BCHR and I’ll achieve it” script}.
You know what? There was an old trick in Oracle 7 you could maybe still look at as a bronze bullet. Put your online redo logs and key index tablespaces on the fastest storage you have and split your indexes/tables/partitions across the faster/slower storage as is fit. Is all your storage the same speed? Go buy some SSD and now it isn’t….

Cartridge with Wooden Bullet

Then there are bronze bullets that you can use that very often improve performance but the impact can be catastrophic {Let’s call them wooden bullets :-) }. Like running your database in noarchivelog mode. That can speed up a lot of things, but if you find yourself in the situation of needing to do a recovery and you last cold backup is not recent enough – catastrophe. A less serious but more common version of this is doing things nologging. “oh, we can just re-do that after a recovery”. Have you done a test recovery that involved that “oh, we can just do it” step? And will you remember it when you have a real recovery situation and the pressure is on? Once you have one of these steps, you often end up with many of them. Will you remember them all?

How many of you have looked at ALTER SYSTEM SET COMMIT_WRITE=’BATCH,NOWAIT’? It could speed up response times and general performance on your busy OLTP system. And go lose you data on crash recovery. Don’t even think about using this one unless you have read up on the feature, tested it, tested it again and then sat and worried about could possibly go wrong for a good while.

That last point is maybe at the core of all these Performance Bronze Bullets. Each of these things may or may not work but you have to understand why and you have to understand what the payback is. What could now take longer or what functionality have I now lost? {hint, it is often recovery or scalability}.

So, what was that one Silver Bullet I tantalizingly left hanging out for all you people to wait for? You are not going to like this…

Look at what your application is doing and look at the very best that your hardware can do. Do you want 10,000 IOPS a second and your storage consists of less than 56 spindles? Forget it, your hardware cannot do it. No matter what you tune or tweak or fiddle with. The one and only Performance Silver Bullet is to look at your system and your hardware configuration and work out what is being asked and what can possibly be delivered. Now you can look at:

  • What is being asked of it. Do you need to do all of that (and that might involve turning some functionality off, if it is a massive drain and does very little to support your business).
  • Are you doing stuff that really is not needed, like management reports that no one has looked at in the last 12 months?
  • Is your system doing a heck of a lot to achieve a remarkably small amount? Like several hundred buffer gets for a single indexed row? That could be a failure to do partition exclusion.
  • Could you do something with physical data positioning to speed things up, like my current blogging obsession with IOTs?
  • You can also look at what part of your hardware is slowing things down. Usually it is spindle count/RAID level, ie something dropping your IOPS. Ignore all sales blurb from vendors and do some real-world tests that match what you app is or wants to do.

It’s hard work but it is possibly the only Silver Bullet out there. Time to roll up our sleeves and get cracking…

{Many Thanks to Kevin Closson for providing all the pictures – except the Silver Bullet, which he only went and identified in his comment!}

IOT part 3 – Significantly Reducing IO

<..IOT1 – the basics
<….IOT2 – Examples and proofs
IOT4 – Boosting Buffer Cache Efficiency..>
IOT5 – Primary Key issues….>

In the previous two posts I covered the basics of Index Organized Tables (IOTs) and then created some test tables to demonstrate the benefit of IOTs that is most often covered – reducing the IO needed to get a single record by one IO, say from 5 to 4. {Whether this is a buffer get from memory or a disc IO depends on if the block is cached, of course}.

In this post I am going to show how IOTs can far more significantly reduce the IO when several related rows are required.

Below is one of my test tables, the one that is a normal heap table and has a primary key, CHHE_PK on PARE_ID and CRE_DATE:

mdw11> desc child_heap
 Name                                      Null?    Type
 ----------------------------------------- -------- --------------
 PARE_ID                                   NOT NULL NUMBER(10)
 CRE_DATE                                  NOT NULL DATE
 VC_1                                      NOT NULL VARCHAR2(100)
 DATE_1                                             DATE
 NUM_1                                              NUMBER(2)
 NUM_2                                              NUMBER(2)

mdw11> select count(*),count(distinct(pare_id)) from child_heap

---------- ------------------------
   1000000                     9999

As you can see, the table has 1 million records and 9,999 values for PARE_ID, there are approx 100 records per parent. The data was created to match a common situation – that of a bit of data coming in for each parent every day. See post 2 for details.

The result of this is that the data for any given parent is scattered through the table. As the data comes in for a given day, the data for the first parent is added to the end of the table, followed by all the data for all the other parents who have data that day. The next day this is repeated, so the child records for a given parent are interspersed with the child records for many other parents.

The below diagram demonstrate what will now happen if you issue a statement like
select *
where PARE_ID=12

Oracle quickly works down the index to the leaf block containing the first key that matches the range. This takes, in my example, 4 block reads. Oracle now works through the index entries and, via the rowid, identifies the exact block to visit in the table for each key. For each key it has to visit a new block – because the data is scattered through the table. This is what the clustering_factor in the index statistics is measuring, how often contiguous rows in the index are for the same block. In our case, almost never.
In my diagram I do not number those table reads but in my simplistic diagram it would be 10 further reads.
If Oracle reaches the end of the leaf block before it reaches the end of the range of key values, oracle follows the pointer in the leaf block (not shown) to the next leaf block (whcih is another block read) and continues working through the keys until the range scan is completed.

In my simplified diagram I only have 6 entries per leaf block. In reality, and in my example tables, this is more like a few hundred. 247 in the case of CHHE_PK.

Now let’s consider my Index Organized Table, CHILD_IOT. It has exactly the same columns as CHILD_HEAP and the data was created in the same way. However, because it is an IOT, as the data came in it was inserted into the primary key index and is thus in an ordered state.

The below diagram demonstrate what will now happen if you issue a statement like
select *
where PARE_ID=12

Oracle works down the index to the leaf block where the range scan begins and now simply works along the leaf blocks. There is no need to go and visit the table as there is no table.

In my IOT diagram the leaf entries are longer and there are fewer in each leaf block, ie 5. So my scan has to visit 3 leaf blocks rather than 2. In reality the difference is more pronounced, in my example table there are actually 56 rows per leaf block, compared to the 247 in the index on the heap table. As such, my scan on the IOT will cover more leaf blocks but this is insignificant compared to the reduction in block visits caused by not having to go hunt down records scattered over the table. Even in the unlikely event of my IOT being deeper by 1 level (an extra layer of branch blocks) due to the reduces entries per leaf block, I would still be winning for range scans.

That is all nice theory and pictures. As ever, we need to back this up with some real tests. Firstly, I am using SQL*Plus and I need to set my arraysize large enough so that I do not introduce extra consistent gets through selecting small sets of rows between client and server. You will need to do the same to see similar results to me.
{I keep meaning to do a dedicated post on arraysize but H.Tonguç YIlmaz has a nice post already on it.}

set arraysize 200
set autotrace on

Now I will select all the records for PARE_ID=10, including a column not in the Primary Key, so that the table needs to be visited. I did this twice to remove the parsing overhead:

select pare_id,cre_date,vc_1
from child_heap
where pare_id =10
order by cre_date

---------- --------- -----------------------------------------------------------------------

82 rows selected.

Execution Plan
| Id  | Operation                   | Name       | Rows  | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT            |            |   100 |  6900 |   103   (0)| 00:00:02 |
|   1 |  TABLE ACCESS BY INDEX ROWID| CHILD_HEAP |   100 |  6900 |   103   (0)| 00:00:02 |
|*  2 |   INDEX RANGE SCAN          | CHHE_PK    |   100 |       |     3   (0)| 00:00:01 |

          0  recursive calls
          0  db block gets
         86  consistent gets
          0  physical reads

82 rows collected and 86 consistent gets. That will be 4 consistent gets to process the index blocks and 82 for the table blocks.

Now let’s repeat that on the IOT:

select pare_id,cre_date,vc_1
from child_IOT
where pare_id =10
order by cre_date
mdw11> /
any key>

---------- --------- ------------------------------------------------------------

108 rows selected.

Execution Plan
| Id  | Operation        | Name    | Rows  | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT |         |   100 |  6900 |     4   (0)| 00:00:01 |
|*  1 |  INDEX RANGE SCAN| CHIO_PK |   100 |  6900 |     4   (0)| 00:00:01 |

          0  recursive calls
          0  db block gets
          6  consistent gets
          0  physical reads

We actually gathered more data, 108 rows compared to 82, all for 6 consistent gets compared to 86 consistent gets. That is a reduction to less than 10% of the original effort.

Now for a more extreme test. I am going to select a single row summary of data for 10 parents, flushing the cache between each run to show the impact when you have to do real IO to support those consistent gets. This is on a fairly old {4 years} laptop with a rather tired hard disc

alter system flush buffer_cache

System altered.

Elapsed: 00:00:00.18


select count(*),sum (num_1)
from child_heap
where pare_id between 50 and 60

---------- ----------
      1155      12031

Elapsed: 00:00:06.39

Execution Plan
| Id  | Operation                    | Name       | Rows  | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT             |            |     1 |     7 |  1203   (0)| 00:00:18 |
|   1 |  SORT AGGREGATE              |            |     1 |     7 |            |          |
|   2 |   TABLE ACCESS BY INDEX ROWID| CHILD_HEAP |  1200 |  8400 |  1203   (0)| 00:00:18 |
|*  3 |    INDEX RANGE SCAN          | CHHE_PK    |  1200 |       |     7   (0)| 00:00:01 |

          0  recursive calls
          0  db block gets
       1157  consistent gets
       1112  physical reads


alter system flush buffer_cache

System altered.

Elapsed: 00:00:00.18


select count(*),sum (num_1)
from child_iot
where pare_id between 50 and 60

---------- ----------
      1111      11528

Elapsed: 00:00:00.29

Execution Plan
| Id  | Operation         | Name    | Rows  | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT  |         |     1 |     7 |    24   (0)| 00:00:01 |
|   1 |  SORT AGGREGATE   |         |     1 |     7 |            |          |
|*  2 |   INDEX RANGE SCAN| CHIO_PK |  1200 |  8400 |    24   (0)| 00:00:01 |

          0  recursive calls
          0  db block gets
         25  consistent gets
         25  physical reads

The Heap took 6.39 seconds and required 1112 physical reads to support 1157 consistent gets.
The IOT took 0.29 seconds and required 25 physical reads to support 25 consistent gets.

I think we can all see that IOTs have the potential to greatly reduce physical and logical IO. Perhaps we should all be using IOTs more.

Final point. The Heap version took less physical reads than consistent gets as some blocks read into the block buffer cache held data required later in the query.

The impact of IOTs on the buffer cache will be the topic of my next post on IOTs. I think { hope:-) } that many of you will be very interested and impressed by what you could gain…