1 00:00:06,440 --> 00:00:11,980 Let's listen we'll talk about how aid is replication works within a site. 2 00:00:11,980 --> 00:00:19,240 We'll talk about connection objects the knowledge consistency checker notification and polling. 3 00:00:19,270 --> 00:00:27,250 These are the concepts relating to internal site replication so into a site replication is a tedious 4 00:00:27,250 --> 00:00:32,140 replication within a single site that the cures automatically. 5 00:00:32,440 --> 00:00:41,560 However you also can configure it to cure manually if application in your environment require a more 6 00:00:41,560 --> 00:00:44,260 specific replication schedule. 7 00:00:44,260 --> 00:00:52,070 As I have mentioned the concepts relating to interim site applications are connection objects. 8 00:00:52,090 --> 00:01:00,040 The knowledge consistency checker notification and Pauline first connection objects a domain controller 9 00:01:00,070 --> 00:01:07,810 that replicates changes from another domain controller is called a replication partner connection objects 10 00:01:08,110 --> 00:01:16,630 link application partners a connection object represents a replication path from one domain controller 11 00:01:16,630 --> 00:01:25,630 to another connection objects are one way representing inbound totally bool replication to view and 12 00:01:25,630 --> 00:01:27,890 configure connection objects. 13 00:01:27,970 --> 00:01:37,180 You have to open active directory sites and services and then select A.D. as certain container of a 14 00:01:37,180 --> 00:01:39,760 domain controller server object. 15 00:01:39,820 --> 00:01:47,740 You can force replication between two domain controllers by right clicking the connection object and 16 00:01:47,740 --> 00:01:49,990 then select and replicate. 17 00:01:49,990 --> 00:01:59,890 Now note that replication is inbound only so if you want to replicate both domain controllers you need 18 00:01:59,890 --> 00:02:05,530 to replicate the inbound connection object of each domain controller. 19 00:02:05,530 --> 00:02:13,510 No the knowledge consistency checker the replication pass that connection objects built between domain 20 00:02:13,510 --> 00:02:22,540 controllers create the forests replication topology you do not have to create the replication topology 21 00:02:22,540 --> 00:02:33,600 manually by default the s create at apology that ensures effective replication the topology is two way. 22 00:02:33,610 --> 00:02:40,080 Which means that even the main controller fails replication continues uninterrupted. 23 00:02:40,180 --> 00:02:49,510 The topology also ensures that no more than three network cops a cure between and it domain controllers 24 00:02:49,720 --> 00:02:58,930 on each domain controller a component of a deed is called The Knowledge consistency checker helps generate 25 00:02:59,020 --> 00:03:07,720 and optimize the replication automatically between domain controllers within aside the knowledge consistency 26 00:03:07,720 --> 00:03:16,450 checker evaluates the domain controller as an aside and then creates connection objects to build the 27 00:03:16,450 --> 00:03:26,020 two way 3 Hope topology that was described earlier if you add or remove a domain controller or if a 28 00:03:26,020 --> 00:03:28,670 domain controller is not responsive. 29 00:03:28,810 --> 00:03:37,210 The knowledge consistency checker rearranges the topology dynamically adding and deleting connection 30 00:03:37,210 --> 00:03:46,330 objects to rebuilt an effective replication topology the knowledge consistency checker runs at intervals 31 00:03:46,630 --> 00:03:55,990 that you specify and runs every 15 minutes by default it designates a replication roads between domain 32 00:03:55,990 --> 00:04:04,240 controllers that are the most favorable connections available at the time you can create connection 33 00:04:04,240 --> 00:04:12,760 objects manually to specify a replication path that should persist however create and a connection object 34 00:04:12,760 --> 00:04:20,980 manner that typically is not required nor it is recommended because the knowledge consistency checker 35 00:04:21,010 --> 00:04:29,330 does not verify or use the manual connection object for fe lower their knowledge consistency checker 36 00:04:29,350 --> 00:04:36,790 also will not remove manual connection objects which means that you must remember to delete connection 37 00:04:36,790 --> 00:04:39,670 objects that you create manually. 38 00:04:39,670 --> 00:04:45,030 The next concept of intercept replication is notification. 39 00:04:45,160 --> 00:04:52,450 So when a change occurs eurozone an active directory of partition on a domain controller the domain 40 00:04:52,450 --> 00:05:02,440 controller use the change for replication to its partners by default the source server waste waits 15 41 00:05:02,440 --> 00:05:11,860 seconds to notify its first group location partner of the change notification is a process by which 42 00:05:11,860 --> 00:05:19,720 an upstream partner informs its downstream partners that a change is available. 43 00:05:19,720 --> 00:05:26,920 By default the source domain controller then waits three seconds between notifications to additional 44 00:05:26,920 --> 00:05:36,530 partners these delays or the initial notification delay and the subsequent notification delay Star or 45 00:05:36,550 --> 00:05:41,190 the network traffic that entire site replication can cause. 46 00:05:41,260 --> 00:05:48,730 After receiving the notification the downstream partner requests the changes from the source domain 47 00:05:48,730 --> 00:05:56,320 controller and the director or replication agent pulls the changes from the source domain controller. 48 00:05:56,410 --> 00:06:06,970 For example supposed domain controller DC 1 initiates a change to a tedious when DC to receive the change 49 00:06:06,970 --> 00:06:08,210 from DC 1. 50 00:06:08,410 --> 00:06:17,080 It makes the change to its director DC to learn to use the change for application to its own downstream 51 00:06:17,080 --> 00:06:18,220 partners. 52 00:06:18,220 --> 00:06:24,800 Next suppose this is 3 is a downstream replication partner of DC 2. 53 00:06:25,090 --> 00:06:34,930 After 15 seconds the city notifies DS 3 that it has a change DC 3 makes the replicator change to its 54 00:06:34,930 --> 00:06:39,730 director and then notifies its downstream partner. 55 00:06:39,730 --> 00:06:48,040 The change has made two hops from DC one to this two and then from DC 2 to DC 3. 56 00:06:48,040 --> 00:06:56,270 The replication topology ensures that no more than three hops a cure before all domain controllers in 57 00:06:56,280 --> 00:07:04,930 the site receive the change at approximately 15 seconds or hope the change fully replicates in the site 58 00:07:04,960 --> 00:07:11,130 within one minute and the last concept of interim site replication is bawling. 59 00:07:11,140 --> 00:07:19,080 Sometimes a domain controller might not make any changes to its replicas for an extended time particular 60 00:07:19,080 --> 00:07:20,980 or duration of hours. 61 00:07:20,980 --> 00:07:24,040 Suppose this is the case with DC 1. 62 00:07:24,040 --> 00:07:32,200 This means that DC to its downstream replication partner will not receive notifications from DC one 63 00:07:32,430 --> 00:07:39,970 DC one also might be off line which would prevent it from sending notifications to DC 2. 64 00:07:40,060 --> 00:07:49,520 It's important for DC to to know that its upstream partner is on line and simply doesn't have any changes. 65 00:07:49,570 --> 00:07:53,850 This week yours through a process called Poland during Poland. 66 00:07:53,870 --> 00:08:03,520 The downstream replication partner contacts the upstream replication partner with Waris as to whether 67 00:08:03,580 --> 00:08:07,290 any changes are queued for replication. 68 00:08:07,450 --> 00:08:14,350 By default the Poland interval for in drug site replication is once per hour. 69 00:08:14,500 --> 00:08:23,090 You can configure the Poland frequency from a connection object properties by clicking change schedule. 70 00:08:23,100 --> 00:08:31,270 Although it's not recommended even upstream partner fails to respond to repeated Poland whereas the 71 00:08:31,270 --> 00:08:39,520 downstream in the Downstream partner launches the knowledge consistency checker to check the replication 72 00:08:39,520 --> 00:08:40,480 topology. 73 00:08:40,660 --> 00:08:48,970 Even the upstream Sora is indeed off line their knowledge consistency checker rearranges the site's 74 00:08:48,970 --> 00:08:53,940 replication topology to accommodate the change. 75 00:08:53,960 --> 00:09:03,160 Now I'd like to recap this lesson as there are lots of figures here and important time information. 76 00:09:03,160 --> 00:09:09,760 But before the recap I'd like you to answer the following question please describe the circumstances 77 00:09:10,060 --> 00:09:17,090 that result when you create a connection object manually between the main controllers within a site. 78 00:09:17,170 --> 00:09:25,120 Press pause here and think about the answer on this question and the answer is it's not recommended 79 00:09:25,210 --> 00:09:32,920 that you create a connection object manually nor is it required typically because the knowledge consistency 80 00:09:32,920 --> 00:09:41,650 checker does not verified or use the manual connection object for fail over the knowledge consistency 81 00:09:41,650 --> 00:09:49,690 checker also does not remove manual connection objects which means that you must remember to delete 82 00:09:49,990 --> 00:09:53,130 connection objects that you create manually. 83 00:09:53,170 --> 00:10:00,640 Now in this lesson we've discussed the role of knowledge consistency checker in grade and connection 84 00:10:00,640 --> 00:10:08,960 objects to create an efficient three hope maximum and robust two way topology. 85 00:10:09,010 --> 00:10:16,910 I'd like to emphasize that there are few reasons to create connection objects manually within a site. 86 00:10:16,980 --> 00:10:25,410 In fact administrators have very few options by which they can modify the replication topology within 87 00:10:25,410 --> 00:10:26,200 a site. 88 00:10:26,310 --> 00:10:33,690 I'd like to mention that within a single site the replication goal is to update all domain controllers 89 00:10:34,050 --> 00:10:36,120 as quickly as possible. 90 00:10:36,120 --> 00:10:44,430 However when I change a Eurozone a domain controller the domain controller waits as long as 15 seconds 91 00:10:44,730 --> 00:10:47,970 to notify its partners of the change. 92 00:10:47,970 --> 00:10:51,570 This increases the efficiency of replication. 93 00:10:51,690 --> 00:10:55,560 If additional changes occur you're on the partition. 94 00:10:55,590 --> 00:11:00,960 I'd like to point out that with a maximum of 15 seconds. 95 00:11:00,960 --> 00:11:11,250 This means that on average changes replicate every 7 dot 5 seconds a maximum of 3 hopes means that within 96 00:11:11,550 --> 00:11:20,610 45 seconds or an average of 20 2.5 seconds the entire society is updated with a change. 97 00:11:20,610 --> 00:11:28,140 Also I'd like to point out that all partitions that replicate between two domain controllers on a connection 98 00:11:28,140 --> 00:11:31,310 object replicate simultaneously. 99 00:11:31,350 --> 00:11:35,580 There is no way to time the partition differently. 100 00:11:35,580 --> 00:11:43,380 The replication traffic is not compressed because it is assumed that all domain controllers in the same 101 00:11:43,380 --> 00:11:49,620 side will be connected with a fast network connection with abundant available bandwidth.