Lesson 06 Post Installation Tasks, Discovery, and Agent Deployment

hello welcome to Train Signal this is the post installation tasks discovery and agent deployment lesson in the sister center operations manager 2012 training course so what really happens post-installation well up to this point you've prepared your operations manager environment for installation we've installed Operations Manager and we've explored the console a little bit along with the installation of operations manager a few things happened first of all some default what are called management packs were installed and you'll it about management packs later in this course but at the same time operations managers deployed its entire framework around

which the product operates what it didn't do however was deploy software on all the servers which you'd like to bring under the operations manager umbrella and that's what we're gonna start working on in this lesson we're gonna basically extend the operations manager tentacles so that they embrace all of the services all of the servers that we want to have under operations manager of monitoring control but there are some post installation best practices too but you should adhere first of all if you've just finished installing an Operations Manager give the product a few hours to get

its bearings this actually does take some time and I've found that with operations manager truly patience is a virtue if you're impatient with operations manager it's gonna be really tough for things to work the way they're supposed to work because here's what happens say for instance you deploy an agent to a server an agent is the piece of software that communicates with operations manager on some servers that can take 15 20 minutes or even up to an hour for that agent to check back in if you're sitting there hitting the refresh button every 30 seconds

waiting for that agents report back you may be sorely disappointed and you may get frustrated and you may end up trying to troubleshoot things when there's really no problem that's just a way that operations manager operates if you can run the console from something other than one of your management servers the console can be a burden on the server so basically you can slow things down a little bit and impact overall operations manager performance for this course we're going to ignore that best practice and run the console from the server and before you get going

with too many post installation tasks it's good to check to make sure that everything that's been up done up to this point is actually working did operations manager actually install the way we intended it to install when it comes to deciding what you're gonna bring under the operations manager umbrella there's some things that have to happen beforehand really what we've talked about is bringing the entire environment under the umbrella but you may not want to do that there may be only some assets that you really want to bring into operations manager for monitoring first of

all you need to identify those assets that are gonna require business-critical monitoring this may include one two or all of your servers and it may include some or all of your network devices next you have to find a way to tell operations manager how to discover these devices this is a discovery process that we'll be going through in this lesson and next for some of the device types you want to break into operations manager such as Windows servers Linux servers or UNIX servers you need to install an Operations Manager agents on the discovered assets and

once you install that agent this really enables all of the operations manager functionality that we expect you know the raising of alerts monitors being adhered to for making sure that certain system parameters stay within range as we've predefined all of the goodies that are defined an Operations Manager happened because we install it agents to our machines now not everything's gonna get an agent your network devices will not get an agent but operations manager will still be able to work with them but not quite as deeply as you would get from a item with an agent

now that said it doesn't mean that you can't get a whole lot of information about your network devices from operations manager in this course you will be learning how to manage your network with operations manager what are the key items that we did during installation was defined an initial run a secound specifically the management service action account we want to make sure that this has properly been created because this is a critical account when it comes to discovery and for running the agent installation process on all of the machines we discover in this lesson so

as our first demonstration in this lesson we're gonna take a quick peek at the default run as it counts that we're install installation time it makes sure that our management service action account is properly configured and here we are at the console and you can see that I've already gone to the administration area and that's for a reason because this is where we're gonna find our run as accounts if we scroll down the page here we're gonna see that we have a section called run as configuration here we have accounts profiles in UNIX slash Linux

accounts now accounts are just what they sound like they're individual accounts that are used for various purposes in operations manager in fact here's the actual accounts that we specified at the same time that we installed Operations Manager but you could also see that there are other specialty accounts that are used for various purposes for example data warehouse report deployment and let me have network service and local system accounts these are just accounts that are built into Windows you don't really have to worry too much about those particular accounts and operations manager where they never cut

profiles that can tell you how these accounts might be used if we go to the default action account area and I'll double click this you can see this starts a wizard and I'll go to the general properties page it will tell you what this particular profile is used for and you can associate a profile with actions later on in operations manager this is the default action accounts and here's the account that's actually used for that purpose and as you would expect that's the action account that we created earlier on in this course unless you start

getting into UNIX and Linux there'll be a separate set of credentials that will be necessary in order to be able to connect to those systems we won't worry about that quite yet however this is not all you'll see of these accounts as you add management packs to operations manager you won't see new account profiles and new accounts be created and as you install those management packs the documentation will explain to you how you populate this accounts with the appropriate credentials next up you're going to learn how to discover resources on your network that you can

bring under operations managers management umbrella and then learn how to deploy agents to those resources so that you can manage them and monitor them with operations manager without further ado let's move on first of all there are some different agent installation methods the most recommended method is to push installation method in fact it's highly recommended for a number of reasons and I'll talk about those reasons in just a second this push installation method will work on Windows machines UNIX machines as well as Linux machines and here's why it's so awesome it enables automatic ancient upgrades

and console based management of the agents if you push the agent to a machine from the operations manager console that agent is basically administrative lead to the operations manager console the agent itself so what happens is when you have a new version of the agent it just gets upgraded automatically by operations manager however in order for this to work target systems must first be discovered by operations manager now here's the nice thing to know about the discovery process one of the things you may wonder is well if I continually have to discover the same resources

over and over why not just be kind of adding agents all over the place and adding the multiple times and things like that the answer is no Operations Manager has some brains in fact the discovery process will skip machines that are already being monitored so you can basically use that the push installation method it's sort of a Delta based way to deploy the agent so as you add new servers to your environment you simply run the discovery process our operations manager finds anything that's changed since the last time you ran the discovery process and you

then push the agent out to those new resources really couldn't be much easier this makes it really really easy to add operations manage your agents to newly deployed servers once you discover a machine using the operations manager discovery method you can push the agent to that machine now as a demonstration we're going to discover some clients and we're gonna push the agent to a number of machines in the environment for global man --tx once again I'm back in the administration area and that's because this is where we deploy clients from if I go to device

management I go to agent managed right click that and you're gonna get a pretty substantial menu the first item of which is discovery wizard and this is what we're really interested in for this part of the lesson here you can see that you can choose from one of three different kinds of device types to discover in operations manager we are interested in Windows computers and I think next and we can choose automatic computer discovery or advanced discovery which allows us to specify some additional settings for our purposes right now we're gonna do advanced discovery I'm

actually gonna browse for computer names that I'm interested in finding now I could also just choose to scan Active Directory and that's good if you're gonna do things in math if you're gonna do things one or two at a time you can actually just type in computer names and things will work just fine so I'll choose browse I'm gonna do our domain controller and you're gonna see that it shows up as such click the next button and I'll just use the selected management server action account that we've already established and click discover now this process

can take a little bit of time or a whole lot of time depending on the number of resources that need to be found out on the network and it didn't take that much time for us because we're looking for one resource and it's a very small network so I'm gonna choose a device that I want to manage click the checkbox and then you're gonna see here that we have the management server that's associated with this resource as well as the management mode agent or agent lists with an agent you can do all kinds of things

on the machine because the basically the agent has full control of the computer in agentless mode you can't do quite as much and agentless communication is - really burdensome on the management server so just be careful about how many agentless machines you have out there on the network and we'll simply click Next you can see here that I can specify the directory into which I'd like to install the agents there's no installer into the default program files area system center operations manager and we're going to allow the age of to use the local system accounts

when it needs to perform specific actions on the machine and then click finish you're gonna see here that this kind of sends it off to a task area and this allows you to go do other things while operations manager is attempting to deploy the clients of the agent out to the machine that you just specified you can safely close this window at this point if you like but for our purposes here I like to wait until the client is fully deployed so you can see that there's a success message at the end of this process

however that can take a few minutes so I'm gonna pause the video while this process completes as you can see as hoped the tasks completed with absolute success it looks like the close button now right now you're gonna see that this set that's not monitored well that's because it takes a little bit of time for the initial monitoring to begin once you've fully deployed an agent to a machine I can't guarantee you that within 15 minutes to a couple of hours 99% of these not monitored situations go away some take a little bit longer but

these non monitor situations are perfectly normal and basically says right now I'm not sure what's going on with this client I know it's got an agent's but I'm waiting for that agent to give me some initial reporting so right now I'm really not monitoring that guy so don't worry about those effects when we come back later on you will see that that will change to a healthy State now we just got through talking about the automated method for deployment agent to a managed machine and app for operations managers use however you may have to install

something manually perhaps for whatever reason the automated installation isn't working this is behind a firewall or what have you you can also install this as the operations manager in a manual method but there are some downsides first of all the most important downside is agent updates are not automatic so as I mentioned before with the deployment method the push deployment method when there's a new version of the operations manager agent it'll get the point out to the machines however that's not the case if you install the age of manually so that's a pretty big downside

in my opinion so whenever possible you should always use the discovery and push method if you can't go ahead and use the manual method next you have to make some settings adjustments in operations manager in order to enable the ability to manually install agents think about it if anybody could just willy-nilly install an agent to a machine and then expect it to be brought under operations managers auspices well that could be a security risk because you never know what kind of stuff they're gonna try to install on those machines so the default setting is for

Operations Manager to simply reject new manual agent installations and basically what that means is that if you try to install the agent manually operations manager is gonna say yeah that's not gonna happen here go away you can also choose to review new manual agent installations in what's called a pending management view and this means that you can install the agents but then an administrator has to proactively approve the new agent requests before anything can happen if you're gonna use the manual method that's the one that I recommend or you can auto approve new manually installed

agents which is the least secure of the bunch this means that basically anyone that deploys an agent can make its way into the operations manager hierarchy now our demo here is going to consist of a couple of different parts first of all we're gonna configure that managed manual installation approval options that you saw above and we're gonna set the manual agent installation approval options to make administrators have to review new manual agent installations before they can be brought into the hierarchy and then we're gonna manually install and approve an Operations Manager agent first of all

I'd like to put your attention to the fact that we have a healthy status I promise you that would happen it didn't take 15 minutes it took about three minutes but again I can't take a little bit of time for things to become healthy once that agents deployed but our first order of business was to come down here to settings an under server we're gonna choose security and then we'll go to properties right now we're configured to reject new manual agent installations and again that's the high security setting in operations manager however in many cases

you want to manually install the agent maybe you have a firewall that's preventing that the automatic deployments so you're gonna have to be able to somehow manually install an agent so we'll simply click review new manual agent installations in pending management view we will not however choose automatically approve new manually installed agents as again that is a high risk option from a security perspective then we'll do apply and ok and once we start deploying clients with a manual option they're gonna start showing up here in pending management now our next step is to go over

to the sequel server and install the clients manually as you can see I'm now on the sequel server for operations manager and I've inserted the Operations Manager installation media into the DVD drive and as you can see here to start up menu or the setup menu appears under optional installations I'm going to choose local agent and this is going to kick off the manual agent installation process click Next and I'm just gonna choose the defaults for the most part I am going to specify the minitor group information the name of our management group is scum

om and the management server is scum 12-1 now I want to point something out this does not air a check so make sure you're careful about what server name you specify and don't worry about the port keep the port the same but just understand if you have a firewall between this machine and the operations manager server you may need to make some firewall adjustments click Next we use a local system account for action agents or agent action accounts and then we'll click install this will install the operations manager agent with the settings that we just

specified and we have success so we click finish now we're not done the agents been installed but remember it's not actually approved an Operations Manager yet so I'm gonna go back over to the operations manager console and we're not going to approve this agent and look at that back in the Operations Manager console in pending management we have gone - SQL or waiting approval because of a manual installation so I'm going to select this option and click approve over in the task bar and do I really want to move this to the agent managed area

I do now that the agent managed I'll see two items the healthy one that we've already deployed and then not monitored force comm SQL again this will go healthy after a few minutes and once the agent has set an initial update to the operations manager server so that's how you deploy an agent manually you got to see a little bit of this information already but we're gonna talk about agent status information now so you can get an understanding for what the various icons mean in the operations manager console the one we really want to see

as much as possible is the healthy one and that's a green checkmark inside a circle we like that that means all is well with the world and you can rest easy this one is bad all is not well with the world and you can't sleep easy a red X means there's a critical issue somewhere that needs to be resolved then we have a green circle and this may be open it means status is unknown and that's not necessarily a bad thing it just means that right now operations manager truly doesn't know what the Pitt status

of a particular element is a great unknown means the health alerts are no longer being received by the management server from the agents this is different than a green unknown because with a green unknown some health alerts are still making their way to operations manager where they grant no they're not so that means that something is wrong with communication between operations manager and the manager entity and we also have maintenance mode if a little wrench is next to an item it means that system is in maintenance mode now you may put systems in maintenance mode

if you're planning to do a whole slew of Windows updates or during what's called a maintenance window in your organization during maintenance windows you can do any work you want and not have to report it out to the organization and this allows IT departments to be able to keep their systems current without having to constantly interrupt people with unscheduled work during the week our test now is gonna be to review the current infrastructure status and the game basic understanding for what's being displayed in the organization so let's go over to the global band six operations

manager console and get an understanding for where we sit right now well as you can see we have one green checkmark in one open circle remember a green checkmark is great a green open circle is not necessarily bad it just simply means this that it's this unknown right now but operations manager can still see the system basically health alerts are being received by operations manager if that circle goes gray you've got a problem that means health alerts are no longer being received by the management server from the agent and that's gonna mean that you need

to do some additional troubleshooting but if we go to the monitoring area we can see quickly then we've got two systems in a healthy state and one in an unknown state right now and if we click the number next to two it'll bring up two systems they're gonna bring up our domain controller in the system center operations manager system itself and as we scroll across we can see that we have a combination of mostly healthy and a few not monitored multiple not monitored now you may wonder well we've deployed this two windows why is the

Windows operating system reporting back is not monitored here's why the reason is because we have not yet deployed the windows management pack so Operations Manager no so that's a Windows system but that's all it knows it doesn't know what it's supposed to be looking for I know just getting health alerts from the windows system but it doesn't know what to do with any of that stuff so since I'm getting this stuff but I'm not monitoring it actively once you deploy the right management pack I'll go to a checkmark or to a red red X which

is not what we want to see and what knows we have one system in an unknown status and again that's the sequel server that we just deployed and right now we're not gonna see a whole lot because it's still trying to determine what is actually running on that system and that's one of the jobs of the agents the rules you put out there is a by virtue of installing management packs tell the agent what to look for so once you deploy the sequel server management pack later on then the agent says oh I know how

to find sequel server now and I know that this hat machine happens to be running sequel server therefore I'm gonna start monitoring sequel server by virtue of the fact that we've installed this management pack agents have hearts believe it or not the age of that you deploy out to your operations manager manage servers check in with the operations manager management server every 60 seconds that's called the agent heartbeat that's the Asia heart be interval you can change this you can change this globally so you could say I want all of my agents to check in

every five minutes instead of one minute you can change it on a per server basis so you could say I want this one to check in every 30 seconds and I want this one to check in every five minutes but understand the minimum value is five seconds and the maximum value is 86,400 seconds in other words every machine has to check in at least once a day or as often as every five seconds so as a short demonstration I'm going to show you how you can configure the agent heartbeat interval and we're back and we

are trying to configure the global agent heartbeat so we're going to go to the ancient area and choose properties and here we have the global agent settings so every 60 seconds machines are gonna check in and again we can change that to five seconds up to 86,400 seconds so basically at least once a day this agent has to check in to the operations manager server you can set this on a per server basis as well by going up to a server right-clicking it and choosing properties and then choosing override global agent settings then here you

can have a server check in either more often or less often depending on the situation for example if you've got a really mission-critical machine you may want to check in more often than every 60 seconds if however you've got a machine that's either not as critical or as across a slow wham link you may not want to actually have that traffic traverse the way link every minute so you may choose to have it check in every five minutes for every 10 minutes but you can put that value here and for that particular server you'll have

those heartbeat settings now here's another important piece of information how does operations manager determine when a system is down well it's once it misses a certain number of heartbeats remember a heartbeat is every 60 seconds unless you've changed the interval and every minute operations manager is going to expect to hear from your machine so it's gonna check for the first minute and it won't check in then it's gonna start taking concerned and with the second minute it's gonna go okay hey what's going on here and when the third one hits it's gonna say okay we've

got a problem this system is down and you're gonna get a critical mark a red X next to that system however if you have a machine that's across a particularly slow land link or that for some reason can't check in more often than every three minutes you could change that the heartbeat interval but you may not want to do that either because you may want it to be able to communicate with the operations manager management server as often as it can but you can change the heartbeat interval to something else if you like or the

number of heartbeats is something else if you like and again if you're across a slow wind link or something like that you might want to change this to five or even ten heartbeats but generally only if it's the less critical server for your more critical servers you want to make sure that you're being reasonable and not waiting too long to be notified of an outage so our final demonstration for this lesson we're gonna override the default missed heartbeat interval for a server once again I'm back in the administration area and I've gone to settings under

server notice we have another heartbeat option I can choose properties and there's a number of missed heartbeats that are allowed this is an Operations Manager parameter so you can't change this one on a person / basis so you can change this to one to whatever you want so you can tell it how many heartbeats are allowed to be missed before I say that this server has failed and that's the end of this lesson thank you for listening and I look forward to seeing you in the next lesson