Transcript

Thank you for your interest in Precise for Sybase. The goal of this short video is just to provide a brief overview; whet your appetite. We would invite you to request a meeting to look at the solution a little bit deeper.
Let us go ahead and get started. This solution tracks SQL statement execution once per second. Basically, it paints a picture of what the SQL statements are doing as they execute. The contention between SQL statements for resources will show what resources are in short supply, which will indicate (you know) where the bottlenecks are in your system. We offer real-time, alerting. We have a history, so you can do long-term reporting. And that gives you also baselines and therefore deviation from baselines. Precise allows you to drill down on the top ten. Plus, you can isolate by program, machine, by user. And that allows the DBA to look at things more from an end user or application perspective. We track all execution states, including locking and wait states. So, you can always find the issues with Precise. Let us take a look at the expert view. Now, in this example, we are in the dashboard. And we are looking at just one Sybase instance that is running on [IBM] AIX. And you can get some basic information about what is going on. Over here, we are looking at over time; that is the x-axis. The y-axis is where the SQL statements are spending the time. The dark blue (referring to the key) shows us that we are using the CPU [central processing unit]. In total, all of these execution states are the total execution times of the SQL statements. The next stop is the ‘Current’ tab. The current tab shows us what is happening right now, real-time. Now I can think of this of kind of like a picture. Because this is just what is going on right now. Whereas the activity tab shows us what is going over a longer period of time. And so I can drill in here. I can focus on: “Which databases are the busiest?” “Which time frame?” And so I can automatically adjust the time frame by just highlighting and clicking to zoom. That is one of the nice things of Precise: It allows you to drill in, drill out. So, you can always find a problem. Come in. You can then look at what is happening with the ‘inventory’ database in terms of CPU [central processing unit], storage, any kind of lock, etc. I can then focus on the inventory database. What I have done is I have filtered on just the batches. So, within the ‘inventory’ database these batches are executing. Focusing on the top batch: Precise reveals the SQL statements that it submits and they are presented in resource consumption — descending order. So, it bubbles up to the top where are the top resources consumer. And you can see here the dark blue shows that it is using CPU [central processing unit]. I can come over and reveal the text of that batch. So, I can actually see (you know) I am looping through here. I am executing these individual SQL statements. I can also present scalability. And what scalability does is it look at the number of executions in red, and it overlays that with the average execution time. Now if the number of executions and the execution time go up simultaneously, that shows that you are not scaling within Sybase. And so in a QA [quality assurance] situation, you definitely want to take a look at that before you rolled out to production. And to investigate a particular SQL statement, what I am able to do is click on the ‘Tune’ button, and that throws me into this workspace, where I can then take a look at execution plan, the tables that are accessed, the indexes and columns. I can get into statistics, (you know): “What is heavy on the system?” I can also locate where a particular step is executing. So, that allows me to really kind of figure out exactly what is happening within an execution plan. There are a number of additional things that I can view within Precise. I can come in, and I can look at well “What are the objects on my system?” One of the things about Precise is that it captures SQL statements. You may haven noticed that one of the tabs is a history. And because we have the SQL statement and the execution plan that connects us to the objects, we can actually do correlation and kind of show what is happening on your system. So, we can drill in. We can take a look at (as an example) the ‘stock’ table and then reveal the SQL statements that go against the ‘stock’ table and present them once in again in descending order. So what is the heaviest SQL statement against the ‘stock’ table? And you might reveal commonality in the ‘where’ clause, which would kind of tell you (you know) “Hey , a particular index might be helpful and address the usage patterns for a majority of the SQL statements.” So, we can talk about that, how to have an ideally optimized system as well. Next up is statistics. Statistics is going to give you a nice overview of all the usage of the Sybase instance. CPU [central processing unit], connections, transactions. And we can then drill in on all of these different components. And come in and actually look at the wait events. All of the wait events are tracked over time. And so that you can basically see how all of these metrics trend. Precise also has alerting. The alerts are divided into individual sets. And the way I think about this, is the status is kind of: Is it available? How is it configured? But then, we also have load. Because typically an idle system does not have much contention. Everything runs very smoothly. But as load increases, then performance degrades. And then, finally, you start to cross (you know) kind of hard lines in the sand. So these alerts work together to allow the DBA [database adminitrator] to know when a problem is developing, and become proactive — or get an early lead on solving a problem. Actually, I think of proactive a little bit differently. I think of it as maintaining a high-performance environment, so you will avoid alerts altogether. Looking at the general behavior, you can see a just a quick description. This will actually show each of the samples. Every one of them did not exceed a warning or critical level. And what general behavior does is, it looks for non-productive wait states. So it is a good way to gauge if the environment is in a healthy execution state, if it is serving the application well. Precise also tracks the infrastructure. Whether you are running on physical or VM [virtual machine], Precise can monitor what is happening for CPU [central processing unit], memory, I/O [input and output]. We have a plug-in for VCenter, which will detect contention. So, if you are not getting CPU [central processing unit] or memory for your particular guest VM [virtual machine], Precise will notify you of that situation. So, in summary: Precise provides 24-by-seven monitoring. We start at the server level. Precise is Always-On. It is tracking the SQL statements, what the execution states are. So, if you have locking or waits, we can tell that. Precise allows you to drill down. So, you can isolate problems from an application or end user perspective. The Precise solution can be expanded to end-to-end. So, if you would like to actually see the end user transactions from click to disk, you can do that with this solution. You saw dashboards; you saw alerts and reporting. And so Precise can provide information for all stakeholders. This was kind of a high-level overview. And once again, we do invite you to take a deeper dive. Thank you very much.