15:00:14 #startmeeting RELENG (2020-07-14) 15:00:14 Meeting started Tue Jul 14 15:00:14 2020 UTC. 15:00:14 This meeting is logged and archived in a public location. 15:00:14 The chair is mboddu. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:14 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:00:14 The meeting name has been set to 'releng_(2020-07-14)' 15:00:14 #meetingname releng 15:00:14 The meeting name has been set to 'releng' 15:00:14 #chair nirik sharkcz pbrobinson pingou mboddu dustymabe ksinny jednorozec 15:00:14 Current chairs: dustymabe jednorozec ksinny mboddu nirik pbrobinson pingou sharkcz 15:00:15 #topic init process 15:00:27 morning 15:01:43 hello 15:03:17 How is it going guys? 15:04:13 hard to tell yet, just sat down. ;) 15:06:18 well, I have just noticed that half of the nodejs stack is going to be probably retired 15:07:25 jednorozec: Due to no maintainer or non buildable? 15:07:37 mboddu, maintainer 15:07:42 nirik: I am not seeing any fires yet, thats good :) 15:07:47 jednorozec: Okay 15:07:53 Well, lets get started 15:07:56 #topic #9577 Fedora Media Writer 4.1.6 released 15:08:04 #link https://pagure.io/releng/issue/9577 15:08:28 nirik: You are up 15:08:35 Any luck with the win cert? 15:09:14 we got it now. 15:09:22 but I haven't done anything further with it. 15:09:55 do we have a sop on this stuff? or do I need to dig thru old irc logs? 15:10:55 ah, we do 15:10:58 nirik: Well, there is a SOP but its for old style 15:11:01 https://docs.pagure.org/releng/sop_fedora_media_writer.html 15:11:03 As in virtual cert 15:11:32 nirik: Did you get a physical copy or ? 15:12:06 I mean physical USB with cert in it or ? 15:12:12 what do you mean? it's in an email from digicert. ;) 15:12:26 yes, they email certs/keys 15:12:32 Huh? 15:12:47 There goes the security :P 15:14:06 well, it's encrypted smtp... but yeah 15:14:19 anyhow, if you like I can get you a copy. 15:14:30 and jednorozec if he likes 15:15:15 nirik: Sure, that would help 15:15:22 Based on https://en.sklep.certum.pl/data-safety/code-signing-certificates/certum-ev-code-sigining.html 15:15:36 "Certum EV Code Signing Suite contains: USB token combining a smart card reader and a Micro-SD card slot, cryptographic smart card, instructions." 15:15:44 That is what we used before 15:16:29 And I thought they are not allowing any distribution of certs using virtual means, its all hardware (USB) distributions 15:16:39 ah, thats all old. 15:17:41 Okay 15:18:34 Anyway... 15:19:11 #topic nirik was able to get a hold of the windows signing cert, he will try to sign the FMW deliverables when he gets a chance 15:19:18 #undo 15:19:18 Removing item from minutes: 15:19:29 #info nirik was able to get a hold of the windows signing cert, he will try to sign the FMW deliverables when he gets a chance 15:19:57 the mac one will be more anoying probibly. 15:22:11 nirik, send the cert my way as well, please. 15:22:27 jednorozec: sure, whats you rh address again? 15:22:34 nirik, thrcka@redhat 15:22:49 nirik: Yeah, I have never done mac signing 15:23:38 mboddu: well, my mini has been in a box for years... hopefully it still works and I can remember any passwords I set on it to login. ;) 15:24:12 nirik: Exactly same problem and I dont even have my mini with me 15:25:05 do either of you have a windows machine to test the exe? 15:26:19 nirik: Nope :( 15:26:26 I do 15:27:05 ok, cool. might ask you to test... or if you all want to work on signing the win one I can work on the mac one. 15:27:06 But I can get it verified (thanks to friends) :D 15:27:42 * jednorozec has no clue about SW signing 15:27:55 jednorozec: https://docs.pagure.org/releng/sop_fedora_media_writer.html might help 15:28:42 nirik, mboddu ok so let me try to sign and test the exe 15:28:56 yeah, I have no idea if that process will work, but it's a place to start. 15:29:32 nirik, so lets split the work. 15:29:57 sure, sounds good to me if you all are willing... 15:30:58 I will never say no to the new opportunity for hair loss! 15:31:02 okay, I have got access to a windows machine 15:31:03 and we can sign/publish one independent of the other too. 15:32:03 jednorozec: How do you think I lost my hair :P 15:35:28 anything else on this one? 15:37:07 nirik: I was searching for the login to mac mini and I couldn't find it 15:37:27 I have to check with Dennis again as he handed it over to me 15:37:55 #topic #9408 Tag ghc into EPEL8 buildroot for ghc bootstrap 15:38:02 #link https://pagure.io/releng/issue/9408 15:38:48 ah yeah, I never got back around to looking at this one after the last round 15:39:15 I changed the centos ones for the rhel ones... 15:39:53 but that's confused koji. perhaps we should delete the external repo and re-add it? with a different name? 15:40:07 * mboddu checking how I set it up 15:43:40 nirik: Whats the change you made? 15:45:46 I moved all the old directories and replaced them with new directories with the red hat rpms. 15:46:09 so I am confused why koji would see any .centos there 15:46:46 or the old hashes. 15:47:06 so it must have them in the db 15:50:38 nirik: Since they are crb builds, you got the builds and replaced the centos rpms with these crb builds? 15:51:30 you got the builds from rhel* 15:51:32 crb? but yes 15:51:59 Okay 15:52:16 I guess removing the external repo and adding it again might work 15:52:32 Shall I try that? 15:52:36 worth trying... you want me to, or you want to? 15:52:55 nirik: I can try it 15:53:10 cool. hopefully that does it. 15:53:14 #info mboddu will remove the external repo and try adding it again that might fix this 15:53:21 #topic Open Floor 15:53:25 Anybody got anything? 15:53:44 I've added some more arm hardware back in... 15:53:52 and rawhide compose last night was... 15:54:11 a bit under 5 hours 15:54:26 hopefully we can get it down some more. 15:54:48 the 2 longest phases are: 15:54:50 Compose phase BUILDINSTALL time: 1:57:15. 15:54:50 Compose phase GATHER time: 2:51:24. 15:55:07 not sure how to make them faster off hand. ;( 15:56:01 nirik: If we can improve caching on gather phase that might help 15:56:23 But yeah, thats a great jump 15:56:41 I am not sure we are using fully the 'reuse the last compose' stuff 15:56:48 I dont think so 15:57:13 PDC was supposed to solve some of it, but thats a bust 15:57:31 we are using some of it, but not gather_allow_reuse (false by default) 15:58:36 not sure the implications there tho 15:58:45 We can look at it and check with lsedlar about it 15:59:08 yeah. Might be worth trying a compose with it and backing it out if it casuses issues. 15:59:27 thats all I had 16:00:02 Thanks nirik 16:00:18 We are sharp on time 16:00:29 Lets close it for today 16:00:33 Thanks everyone for joining 16:00:36 #endmeeting