Weblogic Performance Tuning.raju

[Music] right on it lets you see today you know so the performance tuning generally okay so can you make it more unique and he's getting some nice from you people you can get any daughter can ask me okay so today we level to see performance tuning so generally what the performance tuning are now we can able to tune the performance for troubleshooting all that we've done each set of instances that's got external issues lets you see today so what is the performance tuning on what are the various whichever way how which will now put you in the performance generally performance monitoring is nothing but generally we are going to be collecting all the I know the budget solving chance data and performance they turn the we had need to be analogous and we need again you know a report now performance information or data actually but in the performance tuning given it is not like that we need to be collect and we need to be little did not the chinking step process which patched into the development staging and they know different different outrageous actually during all changes generally we need to do the performance monitoring should be monitor and drop those offer update you know adjustments there is what actually performance tuning so whenever you get any performance issue and it will be connect the performance data and we need to be do the appropriate adjustments so then we can call it as you know you know performance tuning generally you know performance tuning generally it is not a single ship process which is okay number two they in a different different ways so what are the ways which you have actually hear the to prevent from the bottlenecks actually show me how the following waves to tuner you know the magic shuttering strengths WebLogic sure our performance hardware level tuning operating system level tuning network level tuning on WebLogic level tuning and give him their own tuning JDBC level tuning and Jamie's level two all these we can run it as you know different different levels lets you see the first hardware level tuning generally at the hardware level what we need to do tuned generally see before starting the project in the layout before selling the application to the web London servers need to be examined the performances yeah we need to be examining the performance a number of factors generally influence how much the capacity is given for our wheel configurations so how much capacity which is given for our wheel configurations nearly no so we'll need we'll need to in order to support the regulating server you know so in a given application so exactly what we can able to tune or what we can able to you know predetermined anticipated and you know we think actually so the process of determining the what type of RVs and the you know software configuration is required to meet the applications needs eloquently is called the capacity planning so important interview question what is the capacity planning you have you ever involved in civic capacity planning yes I involved into the capacity planning so skip master planning is nothing but process of determining what type of hardware and software requirement of software configuration is required to me the application needs eloquently is called the capacity plan so generally the documents which provides the capacity planning efforts for solutions like you know web browsing Tsarevich most focus on that hardware requirements generally which we cannot change once you choose the hardware requirements so what are the things generally applications need to be expect or to meet the requirements actually the anticipated number of users the number of signs of requests generally the amount of data and its consistency and determining your target CP utilization and gathering the information about you know all the anticipated number of users number of requests number of action acceptable response time and optionally you can able to you know so come to know configurations like first CPU decides and you know speed sufficiently memory and all these all these which we can able to meet expected to meet the application according to the application requirements then only application can do that expected performance see whenever the request comes to the you know operating system generally almost see period I was going to be takes or almond CP there's going to be rising to complete the request actually CP is enemy see whenever the request is coming from that you know client generally what will happen that request make sure CPU does not utilize you know hundred percent actually CPU utilization or something like that response time wine to put stays constant even increase because of work queuing up in the server so generally you know the TP utilization amazed and seventy percent are so max 80% ship utilization which is recommended if it is cost more than eighty percent generally that kind of request and limits to the network Vitamix always it should be a less than 80 percent 20% should be keep tree for running healthy about the ship utilization so that is about how you can able to you know to meet the application requirements those requirements generally then your application under a high load while monitoring the applications server this and as well as the CPU logician there are server a disk and CPU utilization so those two we can able to monitor and we can able to come to another high load while managing a high load when you are passing the request we can monitor generally so those we can able to check by using that table commands ethylene so table command which will give the information about the complete CP utilization that is about in those generally in the hardware level you can able to tune that only right Halliwell configurations this is first level how to tree on the you know blossom set of incentives at the hardware level so once we will complete in how you can able to tune at the operating system level generally you know you can minimize the bottlenecks at the operating system you can minimize the bottlenecks by tuning all those different areas and second level which will able to see how you can able to tune at the network level see II can you assume what we can able to tune it the network level first we can come to know what is the bandwidth to actually bandwidth is nothing but what not the rate of exchanging or the rate of communicating in between the two processes actually generally it is going to be measures bits per second see every WebLogic sorrow which requires the NF bandwidth to handle all client connects from the clients in that case if does not have the regulars it's our own instance to hand in the client connections generally it does not give any performance maximum performance generally there is a big problem actually makes with the network performance is not affected when the supply of resources is unable to keep up the demand demand for resources in the way so then are they today's applications or today's enterprise applications generally on enterprise lower networks are very fast and are now really you know that it cause of performance is well-defined applications than the B so however generally if you find any work with your network administrator to isolate and eliminate those problems generally you know Sohrab blocking sort of the connections it makes to another type generally in your architecture such as client and database connections so generally those requests which mean it will be forward actually to the different different - actually right so generally we can determine whether you have the enough bandwidth is given a deployment you can use the network monitoring tools how you can able to monitor then now we can able to come tonight the network level performance generally by using the networking tools otherwise by providing the networking operating system vendors to see that what type of load is on the network system and you can also use the common operating system tools such as you know netstat and as well as you know net share comment you can use by using these two commands generally we can monitor both the amount of data being transferred across the network by taking the data transferred in between the application and as well as the application servers between that application server and data server so by using these two or you know operating system no commands or networking networking pools which was able to monitor about that network bandwidth whether it is I only enough or not she does not have enough we decided ticket tolls to the network at much later net1 cut network administrator will come to that part and which is going to be you know resolve that issue so this is the board and rest of things which will do by the network administrator then of the kids will see how we can able to you know to Nate the operating system level generally at the operating system level we cannot able to junior blip parameter so generally what we can able to do we can able to get the following error sexually see whenever we will get any requests from the client the enemy that requests which is forwarded to the operating system first operating from all the requests ones you can able to keep keep in a queue opera which is going to be handling the request what is the wait time and what is not how many requests which we can able to keep in a queue that we need to be defined so by you by setting those parameters which you can able to fix maximum issues at the server side what are those actually typically with written the following error condition Snively TCP tuning parameters which are the relator I know which are caused by the operating system a new mic you know whole circuits from the close wait called uncommon others are my connection refused one too many open files on the server outside and address in use a connector on the client side in the most cases generally these errors can be prevented by addressing the TCP rate value and as well as that queue size so go to the unchipped to the root and you can able to you know set these two parameters and by fitting this we can able to fix maximum he shows at the operating system level go to over there and also whenever you will get too many open files you can able to fix that issue by using by increasing the number of open files so by using the you limit commands that will go to here you limit so you'll able to get by default unlimited generally - yay it's not actually you limit so generally number of own Wells one zero to four you limit - again spinning its you limit commands generally which we can able to set the number of open files by using the new limit command whenever you get the too many open files either you can able to close the existing open files otherwise you can increase that you know number of open the files also like since double file if I up to that you can able to increase the maximum number of open points it will be mad one six double 5v5 so in this way you can able to set some of the issues at the networking level and as well as you know operating system level then third part web logic level general let me see what is the JVM level to make generally what is the JVM generally what are the areas was name and how we can able to give my performance at the DVM level if you just go to over there so at the Trivium level lets you see the first what is the JVM and what is the JVM our picture generally see JVM is nothing but what you know so Java Virtual Machine which converts you know machine Porto bytecode elderly which converts mission code to byte code that is called JVM generally JVM is required whenever you are installing the application development server so to compile and running those applications then only in the user can able to access so that is about the JVM so what are the default memory arguments whenever you install the WebLogic server whenever you start the WebLogic server instances Bochum here and you can check it over there so key so these are all the main default memory arguments so whenever you start the server instance in the enemy know that memory how it is going to be divides then now it has going to be utilizing by the JVM generally see the entire JVM generally does going to be divided into the following areas literally hidden and as well several special one and several states you to hold and as well as you know new generations little units you see this new and as well as older generation we can call it at the heap size and the point addition you can tell it as you know non heap size let's we see what is the hidden or new or in generation there will be new RN generation which divides again three parts hidden several ratio one and settled is sugar so that one's true survival ratios one is we can keep always a free because whenever readings may serve when the within we can borrow the memory from the server else ratio one this is the mode you know several space you one and you are in addition these two you can call them you are in generation so the memory which is allocated for newly created objects or you know recently created objects which you can able to allocate the memory for those objects generally what is voltage generation art in your generation generally long-lived objects which allocates the memory for that long blue objects the both of you know new and the ink generation we can call it as steep sides heap is nothing but what we know which was allocating the memory for the currently active eight active threads are active active objects for generations generally which is not part of the heap set which is you know this is outside of the heap size generally what we can able to do in the heap size generally none each one on heap memory so generally pom sage is nothing but what you know it is a non heap memory which is going to be stores Java classes and methods important interview question what are the default memory arguments you know in the WebLogic server these are all the default memory arguments in the AB logic server see for getting the high performance generally to to grow the know or to grow them I bloodied former memory automatically so that is the reason we will able to give X M is equal to XM s starting memory koto maximum memory XM s means starting memory X MX means maximum memory so we need to be give always you know equal for getting the high performance at the JVM level to prevent from the out of memory to prevent from the out of memory issues and otherwise you know so reducing from the time consumption all those which you can able to you know improve by setting the JVM parameters equals under the X M s and X M X these two parameters should be the same value to get the high performance that is how you can able to tune the JVM values are heap size generally you can able to tune there are two methods actually to tune the first one go to the following path and take the backup of the Z domain Uyeda HH on the open the seldom anyway dot sh and go to the following path go to the following path generally you know go to the memory arguments we need to be change C go to here I'm just changing quite one fight well let's just save and exit and the Ramones goofs-a-lot you can change the memory arguments under the under the domain without it searched and I'm just entering it has gone to restarts and see if you notice or not you just go to over there it's a mess and it's a mix both are same value so which is coming into the running mode so once it is coming into the running wall and you can refresh you so this is generally whenever you change any you know heap size under the under the sin domain in whiticus's which is going to be fixed to the every server instance what I need to do how you can able to you know tune for the specific server instance generally only some specific server or instance so for that login to the admin console click on lock and edit button expand the new government to in the same interval question how you can able to tune the memory arguments for specific server instance expand general mid-north click on the server click on any one of the server and you know so click on Cyril start every server have the one tab called Cyril start go to the arguments and you can type it like this you know 512 512 and so click on save enter the username and as well as password and click on server them so this is how generally you can able to you know configure the you know heap size otherwise memory arguments for that individual server instance so this is about another interruption so generally after completing this tuning part about the JVM generally you know so out of memory issues you can able to get not only because of the space issue insufficient memory you even you can able to get objects are open but cannot close in the sense you know annoys your job objects was there generally those are needs of Java objects which we can able to clear or which we can able to you know free those memory by using the garbage collection what is the garbage collection and what are the garbage collection types was there garbage collection miss generally it has going to be connecting the unusual Java objects you know from the JVM and it has going to be cleaning those memory and the free those memory there are two types of garbage collectors or garbage collections which have minor G C and major G C minor GC means it has going to be collecting the new Java objects from the new or Inc generation and major GC which are going to be collecting the unused Java objects from that old generation lets you see generally minor clique minor GC which is you know easy and fast because it happened to be connecting the need to draw objects from that you are in Generation Next we see here three objects are available in the new are exaggeration two are running one is not running in the me whenever you enable the you know garbage collection logs it has going to be free that memory and only running objects only available here like this M reserve follow move to the survival ratio so this is about how you are going to be you know freeing the memory in the - VC and coming to that major decent company it has going to be collecting the unusual Java objects from the whole generation and they generally enough it is slow because it has going to be take every object see like here in the world generation in which six objects were running here was he was running and three ores not running and which has going to be clean those you know honey the Java objects and free that memory so in this way you know the garbage collection will just want to be improving the performance by freeing the unwanted Java objects how you can able to do that there are following mechanisms actually which we have the following DC types garbage collection types default connector so and parallel DC for the ink generation and concurrent DC for the old generation and you know so these are all the GCS those methods which we'll use internally to collecting that but collection loves less we see how you can able to enable the garbage collection loves to clean the unwanted Java objects or to clean that onion Jojo objects that you see so by using the word boost you see protocol which we can able to enable the garbage collection lugs so what I need to do I'm just copying this one so go to the following path how to enable the garbage collection logs import into into the function go to here there are two methods actually first one using the command mode second one using the admin console also go to here click on admin server close to the monitoring and click on performance then you can able to get collect the garbage collect so you can collect the garbage collection from here also this is one method and second method go to the command prompt and you don't go to the set domain in V dot SH ou take the backup before you are editing anything in the Z domain in with attach it then press I go to the Java options you you so go to the job options here you can type the following code so Arbos vc10 DC details and put in GC timestamps xvc or where you want to store that one you know okay that is for designing or for remembering purpose I gave you the gun let's just skate say one day grid and rebounce the server see automatically which creates go to here see my GC VM comes into the running mode so you go to over there and you can open and I can observe that garbage collection logs before after taking the logs this is what they knew Darwin generation this is what new generation and this your palm generation and it sends non keep size it's very can able to before and after the size which has meant to be default section so in this way the log quenches EEP says generally before and after give a total keep and even DC time also generally it is not making more than 5 5 seconds or generally it is not taking the much time to enable the garbage prediction see generally if you have done it I need to Java with the objects then only we can enable the garbage collection nuts yells it has gone to be decrease in the performance this is about the garbage collection measurement so then after that after completing this one generally you know so in this way I can able to tune the you know the others correction loves then after completing that generally you know now see you have completed out to enable the garbage collection laws you next what I need to do so we label to tune at the next what are the trade they actually how we can able to manage that thread down let's I am moving what is it trade actually ladies we are doing with Land Management Willoughby what is it trade generally what are the types of dumps which are available generally how you can able to manage the credit that is nothing but subtask of the process that is first one second one type of the dumps there are following types of dumps which are available important interview questions this is the first one go down second one keep down third one down these are about the different different dumps generally they actually see the first what is the core dump and double you can able to analysis the code them before that lets you can able to see what is the server hang and server crash see server and server cache means generally you know server hang means WebLogic server process which is running but can cannot able to respond this ends you know hang in place for bug observer job also still exists but cannot able to respond and grab subtle crash or server crash means you chimp let's work let's Rajah a person no longer exists that is called server crash important into the question what is in sorrow smash of sorrow and you know sorrow hang in other words we can palatal sorrow to freeze also that is about to serve our patients are all hang what is a pole down generally you know cordon which is not generated by the administrator which is automatically generates weather you know operating system whenever the server crash so whenever the application gets you a binary code file which produced when the register of persons terminates due to some invalid native poor so generally several crashes nerium crashes machine crashes so at that time automatically chosen division it is the co down and which is available under the domain home so what are the reasons actually to crash or to to generate the for dump generally you know there are following reasons native I go the reverse and the type to divers need to libraries and they sell native libraries most of the case of JVM itself only and sometimes that is what will happen it has going to be generated now inches underscore underscore P edit star dot map file by using this last kind of you can able to tell about showing that you can able to get map in that log file is generated so whenever you may have the number of poor whites number of poor friends which is not zero Corphish is ready for zero you can able to change to that you know some thousand then only it has going to be thinner it is not that log frame otherwise it cannot able to you can do it you know you are going to be you know you know troubleshooting the core dogs generally once you'll able to get then you can able to troubleshoot that core dump that is about the core dumps generally you know once you will complete the core dump we're just moving into that heap dump generally so what is the heap dump gentleman he made a very important into the portion heap dump is nothing but what you know so whole dump generally which crashes whenever the forget crash generally what is the heap dump generally keep them is nothing but what you know so very very important one it is a snapshot of the Java process heap memory all right so at a given time it is a snapshot of that Java process Eve memory at a given time so generally it's a it's a useful tool to troubleshooting the memory leaks and as well as you know any other by mother nature issues by using the heap dump how you can able to obtain the you've done this is syntactic sugar to obtain the heap dump so what I need to do so I can able to obtain the job Society instance server ID then you can able to issue this command and which you can able to dump every every he heap into this if drop bin file then after you obtaining the heap dump you can analysis by using the math tool memory analyzer to how you can able to analysis and you can open that you know he brought bin file which we can able to read by using the math tool only in that you can able to come to know is there any memory leaks was there is there any core issue is there any later to the application level everything we can able to identify then you can able to fix according to the requirement if it is nested in eater to the developer generally we can write the ticket towards to the developer and we can able to fix that issue that is about how do I nemesis the heap dump generally what I need to do next go to here how to obtain the hyuga go to the following table on this pot I need to be get the puff steady and next what I need to do two three three three three this is about the puffs ad so dot slash J map [Music] right so you just going to be generating and dumping into the heap got bin file and you can able to obtain the heap dump by using the different different you know methods also like by using the memory analyzer tools called J console and JRM see also you so press control-c the revival will hear you've dumped in and you can analysis by using the mat - so this is about the heap dump generally so next one what is the plane dumps and how you can able to analysis the rhythms very very important into your question so three dump is nothing but it is the textual dump of all active threats see every request which is managed by using the threats only it has going to be textual dump of all active clips and monitor saga all java applications running in a virtual machine that is called thread dump so generally when we will able to take the freedom stimuli there are following situations or there are there are following reasons actually a request or loaded connection pool fields and pain cannot be read right and queues are overloaded and sorrow hang on back and legacy connection issues these are some reasons and to take the freedom sexually what are the tools to analysis the freedoms actually some right javathread dump analyser random viewer Eclipse knock on laser plug-in all these are the different different tools for analyzing the heap dump still early you know so we were able to see by using the samurai tool it's an open source how you can able to analysis by using the thumb write all the part will able to see this is about the thread dump and what are the reasons and you know what are the tools which we can able to analyze the third ups next one of the ways to obtain the three dumps actually totally Erica we have five methods actually the first method using from the admin console go to here you know administration console and click on server and click on server name click on monitoring and click on performance go to the dump thread stacks like these you that explains in alignment nor click on the servers and you know click on solo instance click on monitoring click on performance and click on Tom let's type this is the thread number actually see here and some traits are waiting some plates or runnable some plates are in a block or copied with the dump and paste into the notepad and analysis it that analysis part will able to see in a lost method she's first method how you can able to obtain the thread on and the second method generally using the G stack actually so using the G stack commands in and what we can able to see you can able to obtain the pure dumps go to hear you go under the followings go to the jaw of home bin under this first dot slash the pm's I can able to come to know the server ID and you know dot slash they stab you to double 3 3 greater than slash home slash Linux desktop freedom that window tasty which is greater so we are just you know creator by using that specific way see some pets are waiting runnable sometimes arena so all these different different things which you can able to see next in the last level this is the second method how we are going to be obtaining the dump and what is the third method generally you know so using the using the Java space WebLogic admin utility copy this one generally what I need to do go to here and copying share see WebLogic tree username password and you know six zero eight zero and angles I'm just entering does going to be general data dump see the trend um has been printed and you know that server standard on motor you can copy and paste into the notepad and they can analysis it so this is how you are going to be you know configuring the configuring the thread amp actually so once we will completing this configuration of the thread dumps you know using the fourth method and fifth method for third method using the fourth method generally we can take the freedoms by using the double olicity also other than what i need to do your space logic dot W LSD I'm just centering then after that connect user name WebLogic and password [Music] so what a little do I need to be connect your space right so next connect username WebLogic and password closet 3 and the t3 so once it is connected to the running server instance in a general thread um so see successfully returned to this log files generally what I need to do exit so generally you know read impeach is generated and they can open and you can analysis it this is third method fourth method and fifth method by using the peak in command generally we can come from of that process lady so then after that you can kill that process ID generally it is going to be taken to the rhythm and you know so processing color that you know defaults are allowed file so peak in space iPhone three space PID to double three double 3 so 1 2 3 4 5 6 like that you can take like you know so by using the different different intervals actually this novel which is regarded as a server log so 5 meters which I obtained actually all the drill dumps then once I have turned in the 90 TB analysis that analysis part or we can able to you know down or I can able to to that analysis generally you know so I can able to do that analysis part see generally whenever you open the thread down in the trend of generally which we have the following threads which are available so generally what of drugs JVM threads and region thrilled our default Ruby logic block socket mixture and WebLogic canal system WebLogic dot admin or ma logic that admin and Revlon Chanel default all these similar which are available based on things these debts you can come to know which then I was going to be leads to the deadlock or any other reasons and we can able to kill those threats and they can be both or you stop those traits let me see what are the stressors States which are available in the trade-offs generally there are following states to locker waiting means these tests are either sleeping are waiting for the notifications act or means these threats are either running are ready to Rhonda when they get they turn at the CPU and lock from incidentally these threats are waiting to get a lock on an object or generally what will happen this happens if two or more threats try to execute the same code there is what actually lock eclipse all the magnets that are leads to that deadlocks these are about the pledge statuses you know and what is a deadlock actually that as we see how you can able to analysis with it down in atoms can be analysis by using the Samba I told in the D so assume like that server is considered as a hunk because of thread dump or waiting to acquire the lock an object waiting to be notified of the sub conditions it'll be seeing summarize the tool to analysis the note atoms you know it is the TA tune actually so based on the random tap general is going to be shows visual depiction of the waiting means gray color active means white color and you know or an augur means red color it is going to be shows then then after generally random generally it happened to be you know you know to look about the look of all done you know deadlocks once you can able to find those things and early we can able to close or you can able to be reestablished both its that is about how we are going to be analysis see next we see it will be display like this actually left side so red colored means it is only those are which are going to be log section all the color it is going to explain well done that we can able to publish so if you shoes belongs to the web browser already later we will troubleshoot by the WebLogic server that will set up if you choose belongs to the application developer and we'll density bestowed towards to the application developer and this she'll it belongs to the database developer a database administrator generally will registry comes towards to them back in inspection we can able to you know and any system that dumps lets you see how we can able to analysis generally i already downloaded some by tool to my local machine family so go to that like some right out there which is available here Java's faced iphone huge arms based on guide our job I just Center develop in the welcome screen will open that browse the log Quinn fine so browser thread um file open so go to the desktop I already have here thread dump dot it's in admin so click on okay see it will display here and if you click on the thread um type in the lane you know today I was going to be displaying this not have seen right similarly which you can able to see here this one actually is a dump tape and it will display all the nodes here what are the threads actually you know see it has going to explain with the dump ABS you know you sequence and table so it will be it would be like these sexual you can able to click on that and you can able to Monday to this color means nearly running red color means block and blocking Eddie and absent and these all they actually did not it's so based on this you can able to get come to know and a troubleshooting this is how you are going to be analysis using that some right tone so this is about that then dumps and are we can able to analysis and we can able to you know troubleshooting they shoes whenever you will get the thread later and more related issues then after completing that next what I need to do how to analyses the memories in the layout to monitor the jvm otherwise Oh Anna memory unless the troops there are following memory unless the probes never mean so if you are starting in a domain in the development mode you know how the G are given William otherwise the console if you are in production mode generally which one how that they are them city rocket machine control those of the three different tools which you can able to monitor their you know memory analyzer memory about the JVM slenderly these are generally default which you can able to get if you want to can go with the third-party tools also lets you see for that go to the following path go to the java phone bin under this door she loves you one VM the moment a click on it will open the following [Music] right it will open like this generally see a robot actually local if you are connecting to the remote system at you let's play here even you can able to come from the code room from here also click on the start page generally what we can able to see click on the visual virtual machine VM generally and this is what actually it will display and click on here and click on this this is a word generally you know logic server memory generally what it has meant to be taking what is the say of data memory also means system properties if you have any Eve dumpster a dumps and profiler steps everything which we will able to discuss here they are displayed here visual VM so it I was going to put in the displaying complete information about the virtual memory so complete information and the click on web logic gates then it has not explained the valta logic and click on monitor you can monitor about that you know so generally logic server uptime along this is Sara's going to be running then you know what is the CPU load average then the day or two CPUs is what is the tea soup DC activity and you know part of a class of total class is loaded and total shared lessons loaded how much memory T is going to be available here the heap size as well as you know used non use leaf size and form generation if you click on the palm generation it is going to be spring used and left one generation space and the traits or managers are going to be live and demanded spent system threads and click on pounds you see and you can able to get the DC here and click on the heap dump you can able to get the give company one from that II from the jewelry and copy and paste into that and you can another see see this is how you are going to be you know from monitoring the monitoring about the Blood Service JVM steadily click on the lips she builds it based on this you know great the statuses which you can able to monitor both you know running sleeping and evaluating and monitoring bits so if you click on over there it will explain I did itself innocent it's the monitor from this lightly sampler means you can monitor about the CPU memory almost memory - going to be consuming or much still it is left-click on profiler you can give the performance about the application generally so in this way you can able to get about that you know complete information about the game is your medium by using the name is Dylan VM - which is only you know region monitor only the Sun JDK not for the jet rocket Italy and it's but seen Sun JDK only you show now the other tool called being console so click on here see click on connect see you can able to monitor about the heat plates and of the class and CPU says and clip on the memory you can monitor about the memory almost still it used generally how much it is committed and how much still it is you know this is the time was there everything which you can able to monitor here click on the threads generally you can able to take the thread um from here also you can detect the deadlock then if you have any deadlocks and classrooms you can monitor both the classes and the virtual memory summary here completely and the EEM means you can able to monitor about in the form of ambience in this way you can able to kinases and monitor about the logics roj and slenderly WebLogic server geum's never knew can monitor like this so this is about how we are going to be monitoring and of analysis in the JVM about the Sun JDK similarly for J ROC continually whenever you start the anon JVM in the production environment dot slash DRM see so it's a very very gentle in the production environment to do detecting the you know memory leaks and as well as you know monitoring about this you know even performance and as legend it can record about the JVM performance all those information generally which you can able to do by using these we are MC j-rock admission control in the production environment to see diddly which we can able to do see what we can able to see here there are technician control generally which controls a lot of pool generally to managing monitoring profiling on the troubleshooting your java applications there are technician controls in wisconsin stock management console like you know like you know you can manage and monitor than Java applications on their iPad TV on time and you can record the you know behavior of the j-rock III we might run times in the day analysis and then she's and you know recording off frame by using the day rocket flight recorder tube the memory leak detectors nearly it has money by identifying on top of our applications so these are the things which you can able to do by using that you know j-rock admission control next what I need to do after completing these I can able to switch to the Dino click on this and you can able to monitor about this so what I in this way you can able to monitor about now logic control generally you know there are zero cognition control so it will be open that zero cognition control right so you can do that part then you know what I need to do click on here in this clip none gentle open start moving lick on all those and we're starting the console see this way I can able to monitor about the Java heap and you know CPU uses and as well as you know one generation processors memory and even you can able to you know gender this is gentle and click on the image and you can monitor about that click on the runtime you can able to monitor both the runtime information about like you know every information about the runtime and click on it once we can monitor both down at once using the day profiler so this is the word how we are going to be monitoring the you know memory unless the tools for zero admission control so this completely completes about that you know how you are able to kunda at JVM level then of that's how you can able to manage the threads actually what are the incoming equations you are getting so what are those generally what are the include cues and what of the work managers in Italy how you can able to you know contribute the work managers that is important interpersonal game you right first we'll discuss theory about see before eight-point X we don't have the work managers which you have the exude use only what is execute cues and what is the purpose of the explode choose the main purpose of the execute cues you know is to handle the requests and it may be internal or external the question later to the WebLogic server so external requests could be you know those requests could be related to that applications they apply to the inside inside the village external otherwise redirecting the request from WebLogic server to database server and those are all connection elect West actually into the liquid it could be either the communicating in between the manager servers are between the manager server to admin server those are all called generally internal and external requests those requests in a domain name to handle by the threads so nothing but a good cures so generally what will happen Institute queues does not have the Sailfish health tuning and first that is for this an event a second digit event in which cannot able to have the cleaner times a request actually which the government which three questions it would post on a basis if you put next that does not have actually those are the two biggest problem section of the exhibit close to or from the problem Germany so they were came into solution called work managers after thorough main point X on Watts which was the big place to work managers with executables there was many to choose what was the work manual Snively work managers which women have the self tuning and also it is used to be by in the work and allocated to this belt on the execution model that takes into the contact min set are defined to improve the performance and as well as you know throughput that is about the work managers whatever components or components which are available in request class gets a request class and response to a request class content request class and the constraints and as well as you know minimum plates on magnum under constraints and capacity constraints these are the different different you know components which are available here in the WebLogic server go to the admin console generally what I need to do here go to here click on the work management's these are all called you know click on new user are called work management components let me see so we can able to configure and what are those actually work managers can be configured either at the domain-driven application level and model even whenever you punctuate the domain Devon which is configured and maintained by the administrator and which is available are those details which is position under the configuration and the application level application application dot XML model level web dot XML so those are the french-german which are going to be reflected scope of the work managers Melanie which could be any one of the following is actually the default manages global of admin and it's applications corporate managers Level II default work managers means these are used to handle the pain management and performing the self-tuning and English work managers nearly most of the applications you know in many situations generally default work managers may be sufficient for the most of the application requirements you know logic server thread handling generally algorithms and all those most of the applications which is enough to hand in that a budget server default off managers it does not end up generally but I will J will go with that global work managers in the sense you know so these work management's how we can able to configure these are the steps actually how are you gonna build upon figure and these changes in which are going to be defined out of that concrete or Excel them let me see generally you know when to use the application you know or work managers in Italy whenever the default work managers does not enough and you know response term goal is required and the minimum that its maximum that constraints is needed to be specified to avoid the surrounded logged in the knee during those conditions will able to use the work manager the first one and then twist class and request class you know expressing it happen to be insures whenever you will get the high priority work is scheduled before the name is Betty what color was important to you no less important work even though I plan to work in submitted off for the lowest reactive work that is what actually request like for the first class and fails say the crush class means generally we can specify the average grade uses time required to cause of the Blitz and respond ten requests Glassman's Melanie a response time request class this type of the reckless class specifies the response time goals in that milliseconds and you know context to the quiz not only does going to be you know assigning the request you know build on the conviction information also it does going to be classes to presenting the request such as current user power current users groups based on that and let's reach you what is the constraint energy there are two types of constraints which are available maximum minimum constraints led a number of threads which are allocated to include the recruits turn total number of requests that can be executed and cube for the blood cell before it has going to be rejecting that request Magnum that constraint first these constraints with limits the number of concurrent threads executing the request from Bob constrained work set by default is unlimited - one minute constraints which it I was going to be guaranteeing the number of prints the server will allocate store picked up the question - away from the deadlocks what is the capacity constraint generally the capacity constraint uses the server to receive the request only when it is reached its capacity by default - financials unlimited so we can able to configure the generally work managers anyone work managers generally follow the same steps on me login to the admin console generally global work managers which are responsible for configuring and monitoring let me shading beta administrator project server admin not into the blank admin control click on lock under edit button and click on work managers and select the organism component and select the target and click connect each angel go to here I will show you one and they sure are something is quite common so click on the word managers and see global off managers click on new select any one of the component response time request class mandala click on that so provide the name and as was the milliseconds in which milliseconds Melanie the server has going to be response and select the target and click on finish button in this way we can able to confuse a map you know different different work managers depends on the requirement this is about how we are going to be managing with that work so once we will complete this one generally what is Stud thread handling what is so most of the interviews they lost the following question what is the difference between making good clothes and work managers then let me say what is instructed and what is the tuning acceptable is that they did nothing but it is a thread that cannot be completed parent work or cannot accept the new work because of unfinished work are deadlocked including himself that means generally so I can make sure this manages the server's existing socket connections determines that which sockets have the incoming request waiting to be cross hofstra that is about the tuning which affects these are about the GBM level to me completely computes about the DVM level tuning she generally after completing that JVM level tuning what I can able to do so how we can able to Gilda getaways or something like that the DBZ level tuning lets you see the JDBC lemon water can able to cure go to here see generally the first point is database 11 we need to be tuned the the following data sources actually what is those you know the first one we can able to tune the connection pool what is engine capacity what is max capacity and what is increment capacity in production for getting the high-performance 515 file and in in develop and more in development mode 515 file in production model 525 5 for getting the high performance generally for getting the maximum performance what we need to do into the path tree is called max capacity himself 15 15 what is in SIL capacity generally the number of physical connections to create when creating the connection pool in the data source by default one and maximal capacity means the number of physical connections that the connection pool is contained and minimum capacity generally the minimum number of physical connections that the connection pool can contain that is one actually you can able to do according to the requirement and most of the cases we will keep initial capacity is called max capacity that is the first parameter we can able to choose and second parameter statement kiichi sides generally what will happen these algorithm which is used for maintaining the prepared statements stored into the statement can see in our humans lost recently used two bits means the fixed and shape and kg size means whenever you increase the size generally the number of prepared colorful statements stored in the catchy and improving the performance data we see first term it will go to the database and including the Pirates and you know next time onwards keep in to the WebLogic server memory and it won't go to the database simulate from the public servers memory it has cultivated yours and improving the performance this is about the third parameter click on that monster generally taste connection on reserve by default which is visible and you can select which enables the WebLogic server to test the connection before it is given to the client it is going to be reducing that time consumption test frequencies generally 120 by default 120 seconds the number of seconds then they WebLogic server instance waits in between the atoms when a testing unusual connections than that means test frequency then after that Domini you know string frequencies in the lead 900 seconds by default the number of seconds to wait before stinking a connection pool that the you know incrementally increase to meet the demands so this is about the stringing frequency these are all the parameters which will enable that unit now JDBC level then after click on the transactions and you can configure about the Anil are very very important into a question what is the NLL logging last resource by using these which is enables non non JDBC connection or non exceeded EVC connections to parts fit in the distributed transaction by using the lost logging with source and it has been to be enables the transaction optimization so these are the parameters also eight or nine parameters which you can able to do that the JDBC level coming to the Jamis level generally you can understand only misses between nature we can able to to the gym is clear on what I need to do the other day you can able to tune about the Machine size and as well as the queue size and what the threshold and quota so generally threshold encoder which specifies Magda before the memory use and the you know machine size the number of passages so this is what we are doing at the DMS level tuning under the demons level tuning the first parameter threshold and quota threshold encoder which is specified maximum for memory use and machine sides and number of messages available to the Jamis server so by default -1 unlimited what is the quota in the knee the maximal number of bytes that can be stored in the Jamie server by default which is unlimited what is the blocking under policy generally FIFO are primitive so FIFA means first in first out and primitive means no no no but all happen it allows smaller send requests to premium prevent the previous a bigger ones when there is sufficient space for smaller machinist and this is what magnimous is this whenever it required according to the requirement you can change the other five parameters you can able to cube and what is session polls and why we need to be conscious reports which are going to be enables an application to pass up the messages concurrently you can configure click on you well the name and as well as you know listen class and click on ok so this is about the JMS level tuning this completely completes about that the projects are worth tuning actually JDBC level and Jamie's level so the other one actually the game is level which you can able to do persist and store generally for storing the message into the purchased and stroke by default you can aim Buddhist or the JDBC stores and fine store on B so the enemy we can select the default store for improving the performance instead of taking the JDBC store it is going to be improving the performance and you can choose the bridges also where you are sending on the similar messages in between the source and target destinations those are the parameters which have been able to tune in the game is level and JDBC level they've completely completes about that WebLogic server tuning undershirt hello hello hello yes understood rate performance tuning No yeah I'm able to hear me are you wanna hear this we see the lost our victim yes the last topic general units we see what is the backup and recovery that part and as well as you know real-time concepts will able to see tomorrow can it requires one hour classes and I will complete tomorrow and make sure you can prepare this weekend I can able to take the mark interview okay tomorrow is the last class for you people okay hello tomorrow I will discuss entire course on tomorrow I will discuss entire real them part including the issues and everything if you have any doubts go through it and they can able talk all the doubts okay on it right so tomorrow is the last class I can able to take all the real-time stuff and I can complete and then after completing that I will forward the no interview portions also to you people okay all right so and Vijay you can check out I sell resume to you also and they can modify according to the equipment and you can send to me and I will able to go through it and if is there anything to add and I can able to suggest and you can add it okay that all it so thank you will be tomorrow at same time most probably we will able to meet you know at 6:30 or six forty around like that okay so if I will get you know of time generally I can able to start normally 6:30 only here you get delay you can able to start this water or like that Sun okay I will mail you okay so most of the cases we can fix 6:30 only right so thank you so incomplete remaining for tomorrow and we will wrap up that a better organization class bye [Music]

Channel: Kommula Raja Reddy

Views: 13,453

Keywords:

Id: nl_NcJrGFPA

Length: 64min 22sec (3862 seconds)

Published: Mon Nov 21 2016