openstackgerrit | Merged openstack/designate master: Start replacing designate legacy jobs https://review.openstack.org/561194 | 00:18 |
---|---|---|
*** trungnv has quit IRC | 00:53 | |
*** trungnv has joined #openstack-dns | 00:54 | |
*** abalutoiu__ has joined #openstack-dns | 01:51 | |
*** abalutoiu_ has quit IRC | 01:54 | |
*** abalutoiu_ has joined #openstack-dns | 01:57 | |
*** abalutoiu__ has quit IRC | 02:00 | |
*** masber has quit IRC | 02:49 | |
*** masber has joined #openstack-dns | 02:52 | |
*** KeithMnemonic has joined #openstack-dns | 03:15 | |
openstackgerrit | Nguyen Van Trung proposed openstack/designate master: Allow to check pep8 along with D000 https://review.openstack.org/567458 | 03:19 |
*** sapcc-bot6 has joined #openstack-dns | 03:40 | |
*** KeithMnemonic has quit IRC | 03:41 | |
*** sapcc-bot5 has quit IRC | 03:41 | |
*** masuberu has joined #openstack-dns | 03:56 | |
*** masber has quit IRC | 04:00 | |
*** masber has joined #openstack-dns | 04:02 | |
*** masuberu has quit IRC | 04:05 | |
openstackgerrit | Nguyen Van Trung proposed openstack/designate master: Allow to check pep8 along with D000 https://review.openstack.org/567458 | 04:38 |
*** penick has quit IRC | 05:53 | |
*** penick has joined #openstack-dns | 05:58 | |
*** MarkBaker has joined #openstack-dns | 07:31 | |
openstackgerrit | inspurericzhang proposed openstack/designate master: Replace Chinese punctuation with English punctuation https://review.openstack.org/569043 | 07:36 |
*** AlexeyAbashkin has joined #openstack-dns | 07:47 | |
*** MarkBaker has quit IRC | 08:12 | |
*** MarkBaker has joined #openstack-dns | 08:26 | |
*** lxkong has joined #openstack-dns | 08:35 | |
*** salmankhan has joined #openstack-dns | 09:02 | |
*** AlexeyAbashkin has quit IRC | 09:53 | |
*** salmankhan has quit IRC | 10:05 | |
*** salmankhan has joined #openstack-dns | 10:18 | |
*** caowei_ has joined #openstack-dns | 10:51 | |
*** caowei has quit IRC | 10:52 | |
*** caowei_ is now known as caowei | 10:53 | |
*** AlexeyAbashkin has joined #openstack-dns | 10:54 | |
*** rfreire has joined #openstack-dns | 11:58 | |
*** diman has joined #openstack-dns | 13:27 | |
rfreire | mugsie, morning Graham! | 13:43 |
rfreire | That ktibi guy that showed up yesterday | 13:43 |
rfreire | how did that end? | 13:44 |
mugsie | I didn;t see anything else :) | 13:46 |
mugsie | you must have fixed it for him \o/ | 13:46 |
rfreire | mugsie, haha yay!!!! | 13:46 |
*** McClymontS has joined #openstack-dns | 13:55 | |
*** ianychoi_ has quit IRC | 13:59 | |
mugsie | frickler: for the new jobs - is there a reason we are adding py36 as voting in check, but not in the gate? | 14:18 |
*** ianychoi has joined #openstack-dns | 14:18 | |
*** KeithMnemonic has joined #openstack-dns | 14:28 | |
*** rfreire has quit IRC | 14:28 | |
frickler | mugsie: when I developed this, bionic was not officially released yet, so not suitable for gate jobs. but you were so fast with merging this yesterday, I only got round to checking the journal collection did work as planned now. ;) | 14:32 |
*** rfreire has joined #openstack-dns | 14:33 | |
frickler | mugsie: I guess we can leave it as is for a week or so and watch things and then add the job to the gate, too | 14:33 |
mugsie | ah, makes sense :) | 14:33 |
mugsie | I was just happy to see this done | 14:33 |
rfreire | mugsie, looky; side question | 14:33 |
rfreire | mugsie, do you do overcommit in your KVM hypervisors? | 14:33 |
mugsie | In some cases, yes | 14:34 |
* mugsie is currently learn how everything is laid out in here | 14:34 | |
rfreire | mugsie, in who do you rely for the memory overflow? ksm or swap? | 14:34 |
mugsie | eh, that is a very valid question that I do not know :/ | 14:34 |
mugsie | I don't think we overcommit memory | 14:34 |
rfreire | mugsie, I learnt that I was doing all wrong | 14:35 |
rfreire | My custy relied in KSM | 14:35 |
rfreire | but ksm has several limitations | 14:35 |
rfreire | including NUMA boundaries - it never cross a domain to merge a page | 14:35 |
rfreire | thus hindering effectiveness | 14:35 |
rfreire | ksm does no miracles | 14:35 |
mugsie | oh, I did not know KSM was based on NUMA boundries | 14:35 |
rfreire | I learnt that the most effective way to do the overcommit is by providing copious amounts of wap | 14:35 |
rfreire | *swap | 14:35 |
* mugsie files that nugget of information away for when I need it | 14:36 | |
rfreire | qemu-kvm keeps track of hot and cold pages (lru) making cold pages candidates to swap | 14:36 |
rfreire | and smartly swaps them off | 14:36 |
rfreire | so let a 1 TB RAM hypervisor | 14:36 |
rfreire | it is no shame to have 200-300 GB of swap space for the hypervisor | 14:37 |
rfreire | the only restriction is that it should be FAST | 14:37 |
rfreire | read, local SSD hdd or high-end storage LUNs or striped mid-end storage LUNs | 14:37 |
rfreire | fyi | 14:37 |
rfreire | :-) | 14:37 |
rfreire | I have just _ABUSED_ my laptop here; I have put two SSDs in stripe mode | 14:38 |
rfreire | Deployed around 60 GB worth of raw instance memory (in a 32 GB laptop) | 14:39 |
rfreire | man | 14:39 |
rfreire | it worked GOOD. | 14:39 |
mugsie | what laptop do you have?!? | 14:39 |
rfreire | W541 | 14:39 |
rfreire | Lenovo | 14:39 |
rfreire | Thinkpad | 14:39 |
mugsie | oh, they are monsters | 14:39 |
rfreire | heck of a nice system sir | 14:39 |
rfreire | think I'll talk about it in the RDO booth | 14:40 |
* mugsie prefers a mad desktop and a lightwight laptop | 14:40 | |
rfreire | I'm on the road 3 times a week | 14:40 |
rfreire | unfortunately, I need a beefy machine | 14:41 |
rfreire | that gave me serious back ache | 14:41 |
rfreire | which was solved when I bought a stroller backpack | 14:41 |
rfreire | Because it is _HEAVY_ | 14:41 |
mugsie | Yeah, that sounds like a good plan | 14:41 |
mugsie | frickler: the only issue I can see with the new logs is the /etc/bind/* files are getting the wrong MIME types, and being downloaded instead of displayed | 14:44 |
mugsie | but having the bind log is awesome | 14:45 |
frickler | mugsie: yeah, we'd need to rename them all to .txt.gz, but I'm not sure it is worth the effort. might be good enough to have the files at all in the rare case one needs to dig deeper | 14:47 |
mugsie | frickler: for sure | 14:48 |
frickler | mugsie: I'm more thinking on how to make that part dynamic when testing other backends | 14:48 |
mugsie | yeah, that would be good - possibly something in the devstack plugin? | 14:48 |
*** McClymontS has quit IRC | 14:49 | |
frickler | mugsie: maybe I can add variables to the role and set those in the job definition, that might be the cleanest solution | 14:50 |
frickler | mugsie: anyway, time to enjoy the afternoon traffic here, bbl | 14:50 |
mugsie | frickler: enjoy | 14:50 |
*** AlexeyAbashkin has quit IRC | 15:12 | |
*** AlexeyAbashkin has joined #openstack-dns | 15:15 | |
*** KeithMnemonic has quit IRC | 15:22 | |
*** MarkBaker has quit IRC | 16:07 | |
*** MarkBaker has joined #openstack-dns | 16:23 | |
*** diman has quit IRC | 16:32 | |
*** salmankhan has quit IRC | 16:36 | |
*** MarkBaker has quit IRC | 16:37 | |
*** MarkBaker has joined #openstack-dns | 16:49 | |
*** AlexeyAbashkin has quit IRC | 17:04 | |
*** ivve has joined #openstack-dns | 17:41 | |
*** salmankhan has joined #openstack-dns | 18:12 | |
*** salmankhan has quit IRC | 18:16 | |
*** lkoranda has quit IRC | 19:30 | |
*** lkoranda has joined #openstack-dns | 19:35 | |
*** lkoranda has quit IRC | 19:42 | |
*** lkoranda has joined #openstack-dns | 19:47 | |
*** rfreire is now known as rfreire|afk | 20:42 | |
*** lkoranda has quit IRC | 21:05 | |
*** lkoranda has joined #openstack-dns | 21:10 | |
*** KeithMnemonic has joined #openstack-dns | 21:39 | |
*** puck has quit IRC | 22:27 | |
*** puck has joined #openstack-dns | 22:31 | |
*** d34dh0r53 has quit IRC | 22:47 | |
*** d34dh0r53 has joined #openstack-dns | 22:47 | |
*** eglute has quit IRC | 22:48 | |
*** eglute has joined #openstack-dns | 22:49 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!