Note - using ASH and the Top Activity screen require the use of the Diagnostics Pack License.
This post was prompted by yet another performance problem identified using
pretty pictures and Active Session History data. Although, as you'll see, some
pretty old-fashioned tools played their part too!
ASH entries only exist for certain SQL*Net events. As usual, I think the
design is very successful as long as you have a basic understanding of how ASH
The events all fall into one of three wait classes - Application, Network
SQL> select wait_class, name from v$event_name where name like 'SQL*Net%' order by 1, 2; WAIT_CLASS NAME -------------------- ---------------------------------------- Application SQL*Net break/reset to client Application SQL*Net break/reset to dblink Idle SQL*Net message from client Idle SQL*Net vector message from client Idle SQL*Net vector message from dblink Network SQL*Net message from dblink Network SQL*Net message to client Network SQL*Net message to dblink Network SQL*Net more data from client Network SQL*Net more data from dblink Network SQL*Net more data to client Network SQL*Net more data to dblink Network SQL*Net vector data from client Network SQL*Net vector data from dblink Network SQL*Net vector data to client Network SQL*Net vector data to dblink 16 rows selected.
Hopefully that immediately dismisses the suggestion I've heard a few times that ASH/Top Activity ignores SQL*Net events. However, there are some events that are classed as Idle here that will not be captured in ASH data and will therefore not appear in the Top Activity screen. (In fact, even if you were to set the _ash_sample_all parameter you would not see these events in the Top Activity screen, even though ASH would contain entries for them. There is an additional filter applied to the data before it's displayed. In any case, I would *not* recommend setting _ash_sample_all except for fun.)
The most contentious of these is probably
SQL*Net message from client. This event indicates that the Oracle server is
waiting for the client to do something. If the client is a user session, Oracle
is waiting for the user to do something - perhaps someone has a sql*plus session open that they're doing nothing with or they are filling in an application screen with data? From an Oracle perspective that session
is Idle and there isn't much we can do to tune it, including the waits would heavily skew the data we're using for our analysis and, because events in the Idle
class are specifically excluded from both ASH data and the OEM performance
pages, such activity won't appear. That seems sensible to me.
However, if the client is actually an application server,
those supposedly idle events can tell us something about end-to-end system
performance. This blog post describes one real life situation where that event
helped me identify the performance problem.
Back to the problem at hand. Sadly, in the midst of a a very chaotic work situation, I didn't manage to grab any of the graphs in question but we had a persistent problem with our critical batch processes spending time waiting on 'SQL*Net more data from client', which showed up very clearly as network class waits.
Now there are a number of reasons why this might be happening. Just a couple off the top of my head
1) There is a performance problem with the application server so Oracle is expecting more data to arrive more quickly than the app server can deliver it.
2) A network problem is affecting the delivery speed.
Although the issue here turned out to be number 2, I suppose the true underlying issue is that this application is insanely 'chatty' and uses very frequent round trips to deliver data. Maybe the application design could be improved first?
Ultimately we enabled Extended Tracing on the relevant sessions and confirmed that, yes, the sessions were spending most of their time waiting for the next batch of data and also how long these waits were. In practise, I'm not sure how much value the trace files added in this particular case. In fact, I think this is an excellent example of the underlying beauty of the sampling approach. As an event became *slower* it appeared *more often* in the samples and we had plenty of examples of how long some of the waits were. (But only some - probably the longest ones.)
How did we solve this mysterious problem? With two very useful tools
- ping/tracert. I simply went on to the app server, pinged the db server and confirmed that the response time (from my memory) was in the range of 70-80 milliseconds - far higher than I would expect. One of the Java developers then ran a trace route to the server which highlighted that network performance was great until we hit the db server. It turned out that one of the network interface cards had auto-negotiated down to 100M/s and this was the root cause - easily fixed. Sometimes old-fashioned and simple tools are all that you need, once you know where you should focus your attention (and ASH is great at focussing attention in the right place).
- The next tool is one of the reasons I wanted to write this post. What made me try ping? Whilst I would have got there eventually, I thought I would try my performance optimisation secret weapon! One just for the database wizards! Something I've been keeping an eye on is Oracle Support's attempt to give people a methodical approach to solving performance issues via the Oracle Performance Diagnostic Guide on My Oracle Support - Note ID 390374.1. There are several guides on that landing page but here is the link to the Slow Performance PDF. (You'll need an account to be able to access this.) This is an example of the kind of advice there. (Note that this is for SQL*Net message from client, but you would tend to get most events together anyway and there's similar advice for more data from client.)
"Cause Identified: Slow network limiting the response time between client and database.
The network is saturated and this is limiting the ability of the client and database to communicate with each other.
1. SQL*Net message from client waits are a large part of the overall time (see the overall summary
2. Array operations are used. This is seen when there are more than 5 rows per execution on average (divide total rows by total execution calls for both recursive and non-recursive calls)
3. The average time for a ping is about equal to twice the average time for a SQL*Net message from client wait and this time is more than a few milliseconds. This indicates that most of the client time is spent in the network."
I've been loathe to mention this document until now because it looks like a work in progress (albeit not updated since the start of 2009), isn't perfect, has gaps and I can imagine a number of my peers taking issue with some of the content. But for something to aid learning if you're not naturally great at solving performance problems, I can think of much worse starting places.
In the end, what could have been a really tricky performance problem (and might have taken a while to notice) was apparent with a brief glance at OEM and we could then follow up by applying the right analysis tools to the problem.
To finish off, when I searched Google for any examples of 'SQL*Net more data from client', one of the first results that cropped up was a chapter from one of my favourite books - 'Optimising Oracle Performance' by Cary Millsap and Jeff Holt. Well worth a read ....
Note - features in this post require the Diagnostics Pack license
Not long after I'd finished the last post, I realised I could reinforce the points I was making with a quick post showing another one of the example tests supplied with Swingbench - the Calling Circle (CC) application. Like the Sales Order Entry application, CC is a mixed read/write test consisting of small transactions. As always, there's more information at Dominic's website.
One of the main differences to the SOE test is that the CC test consumes data so you need to generate a new set of data before each test using the supplied ccwizard utility. I won't show you the entire workflow here but enough to give you a flavour of the process. The utility is the same one used to create the necessary CC schema in the first place but the option I'm looking for here is "Generate Data for Benchmark Run".
I'd already decided that my CC schema is populated with data for 1 million customers when I created it so I just need to specify the number of transactions the next test will be for. I happen to know that on my particular configuration, a 1000 transaction test will take around 5 minutes to run.
I ran the test twice. I've highlighted the first run here in the Top Activity page.
It should be clear that CC suffers significant log file sync waits on my particular test platform, just like the SOE test. Therefore I'll regenerate the test data set, enable asynchronous commits and re-run the test. Here I've highlighted the second test run.
As well as seeing a similar change in the activity profile according to the ASH samples (the log file sync activity has disappeared as has the LGWR System I/O), there's a significant difference to the SOE test. Because this test run is based on a specific workload volume, as defined by the size of the test data, rather than a fixed time period, the second test run completed more quickly than the first run. The activity only fills the 5 minute activity bar partially, rather than the first test which filled the whole bar.
If you test a specific and limited workload volume it is much clearer from the Top Activity page which test is processing transactions more quickly, based on the Time axis. That's why I didn't pick this example the first time - it's too obvious what's going on!
Note - features in this post require the Diagnostics Pack license
With so many potential technical posts in my pile, it was initially difficult to decide where to start again but I figured I should avoid the stats series until I'm back into the swing of things Instead I decided to fulfill a commitment I made to myself (and others, whether they knew about it or not) almost three months ago.
When I gave the evening demo session in the Amis offices I think the 2 hours went pretty well but, as usual with the OEM presentations, I got a little carried away and didn't conclude the demo properly. (This is also the demo I *would* have done at Hotsos last year if the damn thing had worked first time ) It was a shame because as well as showing the neat and useful side of OEM Performance Pages, it also illustrates one of the common pitfalls in interpreting what the graphs are showing you.
I began by running a 4 concurrent user Sales Order Entry (SOE) test using Dominic Giles' Swingbench utility. I won't got into the details of the SOE test because I don't think it's particularly relevant here but you can always download and/or read about Swingbench for yourself at Dominics website.
I ran the test for a fixed period of 5 minutes using no think-time delay.
Using the capability to look at ASH data in the recent past, the OEM Top Activity page looks like this.
- There was a fairly consistent average of 4-5 active sessions over the 5 minutes period and, looking at the Top Sessions panel in the bottom right of the screen, these were four SOE sessions of similar activity levels and the LGWR process.
- The majority of time was spent on User I/O, System I/O and Commit Wait Class activity, with a little CPU.
- Three PL/SQL blocks were responsible for most of the Commit activity.
- The LGWR process was responsible for most of the System I/O activity.
I'll leave it there for now and won't drill down into any more detail.
In terms of optimising the performance of this test, what might I consider doing?
The most important aspect is to optimise the application to reduce the resource consumption to the minimum required to achieve our objectives. There's a whole bunch of User I/O activity that could perhaps be eliminated? But I'm going to ask you to accept the big assumption here that this application has been optimised and that I'm just using a Swingbench test as an illustration of the type of system-wide problem you could see. In that case, my eye is drawn to the Commit activity.
When I'm teaching this stuff, I'm usually deliberately simplistic (at least at the end of the process) and highlight that what I'm interested in 'tuning' is whatever most sessions are waiting on according to the ASH samples this screen uses. I used to explain how I'd look for the biggest areas of colour, drill down into those and identify what's going on. Sadly, I later heard that someone (I think it was JB at Oracle*) had already come up with a nifty acronym for this - COBS. Click on the Big Stuff! One day I will come up with a nifty acronym for something too, but you shouldn't hold your breath waiting.
So, if I click on the big stuff here, I can see that the Commit Class waits are log file sync.
How might I reduce the time that sessions are waiting for log file sync? Here are a few reasons why the test sessions might be waiting on log file sync more often or for longer than I'd like.
- Application design - committing too frequently
- CPU starvation
- Slow I/O to online redo log files
Whether waits are predominantly the result of CPU overload or slow I/O can be determined by looking at the underlying log file parallel write wait times on the LGWR process but that's a bigger subject for another time.
You can look into all of these in more depth - and should - but as this is designed to be a fun demo of the pretty pictures (it used to be 'the USB stick demo'), I'll simply try to eliminate that activity and re-run the same test. Here's how Top Activity looks now.
Oh. Maybe that wasn't what you expected? OK, the LGWR activity has disappeared, but it seems the system is almost as busy as it was before but that the main bottleneck is now User I/O activity. That's often the way, though - you eliminate one bottleneck in a system and it just shows up somewhere else. It must be good to get rid of log file sync waits though, right? User I/O seems like more productive work and I've managed to make the LGWR activity disappear completely.
But then if you were to look at this graph in terms of Average Active Sessions or DB Time or (as it's more likely to be expressed) how big that spike looks, the two tests would look similarly busy from a system-wide perspective. They were but the real question is - busy doing *what*? There's some important information missing here and Swingbench is able to provide it.
Mmmm, so I wonder what that value was for the first run?
Woo-hoo! *That's* what I call tuning a benchmark - processing almost twice the number of transactions in the same 5 minute period.
So it turns out that the sessions in the database *were* just as busy during the second run (not too surprising seeing as the test has no user think time so keeps hammering the database with as many requests as it can handle) but that they were busy doing the more productive work of reading and processing data rather than just waiting for COMMITs to complete.
I raised this issue of DB Time not showing activity details with Graham Wood* at Oracle in relation to a previous blog post. I think he made the point to me that that's why the OEM Performance Home Page is *not* the Top Activity page. If I take a look at that home page, it shows me the same information as the Swingbench results output did, albeit not as clearly
Looking at the Throughput graph, I can see that the second test processesd around double the number of Transactions per second for the same test running on the same system.
To wrap up (and be a little defensive) ...
- Yes, I could have traced one or more sessions and generated a complete and detailed response time profile that should have lead me to the same conclusion.
- Yes, as this is a controlled test environment and I'm the only 'user', AWR/Statspack would have been an even more powerful analysis tool in the right hands.
- The Top Activity page is not the most appropriate tool for this job but it is handy for illustrating concepts.
- Lest I seem a slavish pictures fan, I'm showing how people might misuse or misundersand ASH/Top Activity. In this case, the Home Performance Page is a much better tool because we're looking at system-wide data and not drilling into session or SQL details.
Oh, and what is my Top Secret Magic Silver Bullet Tuning Tip for OLTP-type applications? (only to be used by Advanced Oracle Performance Wizards)
alter system set commit_write='BATCH, NOWAIT';
Done! In fact, why not just use this on all of your systems, just in case people are waiting on log file sync?
(Leaves space below for angry responses and my withering humorous retorts)
* This is not name-dropping, this is giving due credit to the people who really know what they're talking about
In one of my Hotsos Symposium 2010 posts I mentioned that Peter Stalder had plugged some test results from an earlier blog post into Neil Gunther's Universal Scalability Law to see how well the model applied. Peter's posted his slides now and I've added the URL to the comments thread of the original post so people can see another perspective.
He also pointed out a recent blog post discussing similar subjects at Neil Gunther's blog although I must admit I've only had a quick glance at it because I'm up to my eyeballs in mail at the moment