1 00:00:00,210 --> 00:00:01,070 Hello welcome back. 2 00:00:01,080 --> 00:00:03,470 So let's talk about glacier ARCHIVE RETRIEVAL. 3 00:00:03,480 --> 00:00:10,680 So basically you have archive files or archive data that you would like to have again or read it from 4 00:00:11,190 --> 00:00:12,450 glacier for some reason. 5 00:00:12,450 --> 00:00:14,590 You can retrieve parts of an archive. 6 00:00:14,610 --> 00:00:20,070 So if you have an archive that includes hundreds of objects you don't have to retrieve the whole thing 7 00:00:20,100 --> 00:00:22,590 unique you can retrieve only part of that. 8 00:00:22,590 --> 00:00:27,780 It is very common to group multiple objects and compress them using TARP or a ZIP format before archiving 9 00:00:27,780 --> 00:00:27,960 them. 10 00:00:27,960 --> 00:00:34,620 So here we are talking about like how would this file include or how would this archive include a lot 11 00:00:34,620 --> 00:00:36,110 of objects within it. 12 00:00:36,110 --> 00:00:42,060 It is very common to do that that you either turn it or zip it the multiple files and that constitutes 13 00:00:42,090 --> 00:00:48,390 the archive that you load into a glacier retrieving archives is an asynchronous job uploading archives 14 00:00:48,390 --> 00:00:49,160 to Glacier. 15 00:00:49,170 --> 00:00:55,740 Remember when you said that it gets uploaded to multiple facilities then this is synchronous job and 16 00:00:55,980 --> 00:01:01,530 glacier will not return a success message unless it has been copied to all the different facilities 17 00:01:01,530 --> 00:01:03,770 or the all different storage devices. 18 00:01:03,840 --> 00:01:08,490 So synchronous for upload asynchronous for retrieval and it doesn't happen automatically you need to 19 00:01:08,490 --> 00:01:11,820 submit or initiate a request for the retrieval to be initiated. 20 00:01:11,850 --> 00:01:15,330 You cannot use it as counsel for archive jobs retrieval. 21 00:01:15,330 --> 00:01:18,090 So basically you can initiate the restore request. 22 00:01:18,090 --> 00:01:24,620 You can also like I let's say I want to the bulk and I'm going to now to restore petabytes from Glacier. 23 00:01:24,660 --> 00:01:26,490 Do I just check over five minutes. 24 00:01:26,490 --> 00:01:28,530 No you can't configure that. 25 00:01:28,540 --> 00:01:35,340 The servers the glacier is going to have an S send you a notification when the retrieval job has complete 26 00:01:35,480 --> 00:01:36,270 byte ranges. 27 00:01:36,270 --> 00:01:39,480 Very important feature of glacier. 28 00:01:39,480 --> 00:01:47,760 Now let's say when I grouped multiple objects in an archive before uploading that to Glacier I knew 29 00:01:47,760 --> 00:01:51,270 that I have the files in a specific order. 30 00:01:51,270 --> 00:01:57,480 I have tracking of that in my database the client side one that I maintain and I know which object where 31 00:01:57,480 --> 00:01:58,440 and I know the size. 32 00:01:58,440 --> 00:02:02,060 This starts at byte 1 and ends at byte 1000 for instance. 33 00:02:02,070 --> 00:02:07,410 This one starts at 1001 and ends at five thousand four hundred fifty five. 34 00:02:07,500 --> 00:02:09,030 This one starts and so on. 35 00:02:09,030 --> 00:02:10,140 So I know the by strangers. 36 00:02:10,140 --> 00:02:16,530 So if I wanted to have object 1 and object to only and I don't want the rest from the archive. 37 00:02:16,530 --> 00:02:18,560 So after that has been moved here. 38 00:02:18,630 --> 00:02:22,430 Now I want to retrieve on the object one object too so what do I do. 39 00:02:22,440 --> 00:02:29,040 You can initiate a request an H TTP get request an enraged if you get request you would mention that 40 00:02:29,070 --> 00:02:34,280 I need to get from byte 1 all the way to multiples of 1 megabyte. 41 00:02:34,380 --> 00:02:35,070 So you can. 42 00:02:35,070 --> 00:02:42,030 Here have the sizes or you can have the by ranges but you can also have the sizes and say This one is 43 00:02:42,030 --> 00:02:43,610 point five in size. 44 00:02:43,620 --> 00:02:45,540 This one is one point five in size. 45 00:02:45,540 --> 00:02:47,960 This one is 2.5 in size makes. 46 00:02:48,000 --> 00:02:49,310 I'm talking here about megabytes. 47 00:02:49,340 --> 00:02:55,590 OK so if I want object on an object too I would each issue the request and say I want from one by one 48 00:02:55,920 --> 00:03:02,000 to two megabytes so the end has to be either the end of the archive or multiples of one megabyte. 49 00:03:02,130 --> 00:03:07,140 So you can initiate that and what would happen is Malaysia will only bring you back object 1 an object 50 00:03:07,140 --> 00:03:09,360 to the byte range that you have asked for. 51 00:03:09,360 --> 00:03:15,000 So this is a very important feature especially when you group a large large number of objects into one 52 00:03:15,000 --> 00:03:19,700 archive and you only need to access one or two objects of these or few objects. 53 00:03:19,740 --> 00:03:21,500 You don't have to copy the whole thing. 54 00:03:21,510 --> 00:03:27,030 Okay so the byte range is three and glacier support ETP get request would range in each to be header 55 00:03:27,420 --> 00:03:33,690 so the byte range is in the HDP header to identify a specific byte range you may want to retrieve from 56 00:03:33,690 --> 00:03:35,230 glacier archive data. 57 00:03:35,470 --> 00:03:37,870 Alright so this ends the archive. 58 00:03:37,940 --> 00:03:44,760 The glacial archive retrieval and in the next lecture we are going to talk about the archive or glacier 59 00:03:45,450 --> 00:03:49,620 costs so stay tuned and take a break now we'll see you in the next lecture.