[00:19] it looks like the EC2 AMI for hvm-ssd has not been generated in cn-north-one since 2018-11 - sorry to pop out of nowhere with a question, but where would be appropriate to report/ask about this? [00:20] s/cn-north-one/cn-north-1/ [00:21] jamespage: ^^ is ec2 hvm-ssd image problems on cn-north-one something you'd tend to? [00:21] well, correction: for xenial the last was 2018-11, for bionic the last was 2019-01. also it looks like non-hvm AMIs are still being generated regularly [01:58] when my 19.04 server goes down with multiple ssh client windows, I don't get notifications and the shells stay open [02:16] ChmEarl: I don't have any definitive fixes, just some tips and tricks. 1) if the ssh session is frozen/stuck, type ~. [02:16] (including the period) [02:17] 2) you could "pkill sshd" as your user on the server, to close all ssh sessions for your user before shutting down the server [02:17] 3) you could try "sudo shutdown -r +1" [02:18] ChmEarl: you could also set ServerAliveInterval in your ssh clients, but the downside is the sessions will be much more likely to timeout if a router dies [02:22] ploxiln, sarnold thanks === cpaelzer__ is now known as cpaelzer [06:06] Good morning [10:16] sahid: you might just want to pick the patch from https://review.opendev.org/#/c/663294/ [10:16] sahid: its still needed [10:17] sarnold: fraid not [10:18] jamespage: i've kept it https://git.launchpad.net/~sahid-ferdjaoui/ubuntu/+source/neutron/tree/debian/patches/bug1826419.patch?id=6153732fec521d0a3d8044ad59a26acc6d05f083 [10:19] sahid: ah sorry I was confused by the diff of a diff [10:19] sahid: yes that looks OK to me [10:19] ok cool perfect === cpaelzer__ is now known as cpaelzer [14:01] Hello, im about to migrate a website from shared hosting on bluehost to cloud hosting on Linode, the main goal is site load time improvement, does a dedicated cpu plan make a difference here? [14:01] My plan was to get a ubuntu 19.04 server with 16gbs of ram, 320gb SSD, and 8 core cpu but theres another plan offering smaller ssd and more cores [14:05] kevindank: I would not recommend 19.04 for a server. Not unless you plan on upgrading it in 7 months. Stick with LTS. [14:06] leftyfb: 18.04 lts? [14:06] kevindank: yes, that is the latest LTS [14:06] supported for 5 years from release [14:06] okay [14:08] kevindank: no, you don't need a dedicated CPU linode, the standard will do [14:08] TJ-: Thanks [14:09] kevindank: the 2-CPU/4GB is usually enough for most sites; I host several domains on one such that are relatively busy [14:10] coreycb, sahid: do you think we should start stripping out python- binary packages this cycle? we dropped py2 support in the actual openstack projects, so we could work back down the dependency chain now [14:11] jamespage: yeah i think so. just need to be careful with swift deps i think. [14:13] TJ-: That's typically what I use. I have a website thats on bluehost right now that is maxing out ram,cpu on the bluehost standard server so went with the highter ram plan of 16gb [14:14] kevindank: what's eating the memory? database? [14:15] Total CPUs : 16 / 16 Cores Total RAM : 31 GB Real Time Free RAM : 2 GB Real Time RAM Usage : [14:16] Real Time Free RAM : 1 GB Real Time RAM Usage : 29 GB [14:16] im pulling that uses wp server stats, im not sure if thats the ram/cpu of the shared server or our site on that shared server however [14:17] DB is 78mb [14:18] the site is a lawfirm site, theres a bunch of pages and a chat feature but it should not be using 29gbs by itself at all [14:27] kevindank: is that in the Linode, or at the shared hosting? [14:28] kevindank: I'd presume its the shared host and it is over-subscribed [14:44] sahid: neutron 2:13.0.3-0ubuntu1 uploaded to the cosmic unapproved queue. thanks! [14:46] coreycb: thanks === lotuspsychje_ is now known as lotus|celerynuc === RoyK^ is now known as RoyK [19:21] ploxiln: apparently other images are being updated, it's the streams interface that's giving trouble -- eg https://paste.ubuntu.com/p/PM7QMBNm9g/ [19:23] ploxiln: that was generated via aws --profile=china ec2 describe-images --region=cn-north-1 --owners "837727238323" --query 'Images[*] | sort_by(@, &CreationDate)' --- I don't know enough aws to tell you how to create a 'china' profile like that, but I hope it's something you're familiar with or already done [19:23] thanks for looking into it. the one I'm hoping for is 16.04 xenial with type hvm-ssd - I did an explicit search for it and I think that oen is not being generated [19:24] so I did: aws ec2 describe-images --filters Name=name,Values="ubuntu/images/hvm-ssd/ubuntu-xenial-16.04-amd64-*" Name=owner-id,Values=837727238323 --query "Images[*].[CreationDate,ImageId,RootDeviceName]" --output text | sort [19:24] (with cn-north-1 region) [19:25] and got 2018-11-21T07:51:00.000Zami-013ead89472fc7464/dev/sda1 as latest [19:26] proper paste: https://paste.ubuntu.com/p/bNKnN3sj8T/ [19:29] bionic does have a newer hvm-ssd image, but the previous is from February, and the one before that from August 2018, so I just get the impression that general china flakyness makes each upload a roll of the dice [20:49] So I get a CALL......a Dr is having trouble with his linux system. Its running...wait for it.....Ubunti 14.04 [20:50] I almost laughed in the phone [20:52] whats his phone number. I want to laugh at him for reals [20:54] Windows XP I would have laugh but I'm actually impressed a Dr is using Ubuntu :) [20:56] tiz good stuff [21:16] We are taking it, burning it in the deepest darkest corner of hell and making hime a VM under my controll [21:29] sdeziel, it wasnt as much him "useing" it per se, his grad student set it up so he could run a few programs specific to his work [21:29] but ya we are burning it down...hard