Back to Home / #linode / 2020 / 03 / Prev Day | Next Day
#linode IRC Logs for 2020-03-11

---Logopened Wed Mar 11 00:00:32 2020
00:42<linbot>New news from community: Why is Microsoft blocking my emails? <>
00:54-!-NomadJim_ [~Jim@] has joined #linode
00:54-!-NomadJim_ is "Nomad" on #debian #linode
01:01<nate>The error in that message seems a bit more than just mis-configured DNS and stuff
01:01<nate>"part of their network is in the block list" makes it sound like they blocked a range
01:01-!-NomadJim [~Jim@] has quit [Ping timeout: 480 seconds]
01:13<chesty>i guess that was debaser's question ask in irc above if you want to scroll up
01:17<nate>Yeah I remember it earlier, I just wonder if that was really the case though, ie; the lack of PTR and stuff or if microsoft actually did block a range of linode addresses
01:19<chesty>they had all the right records, got blocked after sending 20 odd emails which all ended in the spam folder
01:27<nate>that'd do it too lol
02:12-!-gparent [] has quit [Quit: Leaving]
02:12-!-gparent [] has joined #linode
02:12-!-gparent is "g" on #linode
02:19-!-gparent [] has quit [Quit: Leaving]
02:20-!-gparent [] has joined #linode
02:20-!-gparent is "g" on #linode
02:34-!-gparent [] has quit [Quit: Leaving]
02:34-!-gparent [] has joined #linode
02:34-!-gparent is "g" on #linode
03:30<nate>god damnit my ipv6 buggery is back, stupid LE requests are timing out once more
03:31<nate>I really need to document the fix to this cause I forget every time it happens lol
03:36-!-dsapikas [] has joined #linode
03:36-!-dsapikas is "purple" on #linode
03:37<chesty>um ipv6
03:40<Zr40>joecool: /48 is still reasonable from an resource exhaustion point of view. On the high end, but still reasonable.
03:43<Zr40>for residential connections, that is. /56 is just fine. /64 is tolerable (just like a single IPv4 address is) but it doesn't let you do anything fancy. /56 should be the minimum
03:45<Zr40>for Linodes I agree a /64 per Linode and a /56 with a good reason is fine
03:48<nate>that still requires me remembering how I fixed it the last times first lol
03:49<chesty>search irc logs?
03:51-!-fergtm [] has quit [Ping timeout: 480 seconds]
04:00-!-fergtm [] has joined #linode
04:00-!-fergtm is "Fernando" on #linode
04:41-!-dsapikas [] has quit [Ping timeout: 480 seconds]
04:57-!-dannyAAM [] has quit [Quit: : ZNC 1.6.2 -]
04:58-!-dannyAAM [] has joined #linode
04:58-!-dannyAAM is "Danny" on #linode
05:52-!-hawk [] has joined #linode
05:52-!-hawk is "hawk" on #linode
05:55-!-avril [] has quit []
06:08-!-avril [] has joined #linode
06:08-!-avril is "..." on #tor-bots #moocows #linode
06:14-!-avril [] has quit []
06:40-!-avril [] has joined #linode
06:40-!-avril is "..." on #tor-bots #moocows #linode
08:40-!-andyzwieg103 [] has joined #linode
08:40-!-andyzwieg103 is "azwieg103" on #linode
08:42-!-chonk [] has quit [Quit: POOF!]
08:48-!-chonk [~chonk@2600:3c03::f03c:92ff:feac:d48f] has joined #linode
08:48-!-chonk is "b" on #linode
08:48-!-chonk [~chonk@2600:3c03::f03c:92ff:feac:d48f] has left #linode []
08:51<dubidub>How many addresses is a /64?
08:53<dzho>you mean, more specifically than 2^64?
08:54<dzho>(128 bits total address size - 64 bits masked = 64 bits for specifying addresses in the range)
08:55<nate>Son of a bitch all those IPv6 issues I've griped about for like a year was a god damned bug in cURL ughhhhhhhhhhhhhhhhhhhhhhhhh
08:55<dzho>even when that's used inefficiently, it is still A Lot™
08:55<dzho>nate: congratulations
08:55<dubidub>So a lot of addresses?
08:55<nate>dubidub: Talking IPv6? A lot a lot yes
08:56<dzho>for large values of "a lot"
08:56<dubidub>Why, then does some Linodes need even more, like a /56?
08:57<nate>dzho: Well it's hard to trace it down to cURL since everything uses it lol. I just found someone on LE forums back in September griping about my exact same issue w/ w/ the issue process taking like 30+ minutes because IPv6 was crawling at like 1 byte/s
08:57<nate>dubidub: It's worth noting the IPv6 assignment you get from linode is not -per- linode, it's per-location
08:57<nate>shared between all linodes you have at that location
08:57<dubidub>ah, right
08:58<rsdehart>so obviously you maximize your address allocation by putting all your linodes at as many locations as possible
08:58<dzho>so if you have an IPv4-sized number of linodes, each will "only" get its own IPv4-sized set of addresses
08:58<dzho>at that location
08:59-!-Madmax [~oftc-webi@] has joined #linode
08:59-!-Madmax is "OFTC WebIRC Client" on #linode
08:59<dzho>nate: so this is a known bug in cURL and a fix is in the latest versions, or pending, or ... ?
08:59<dzho>bug report URL or it didn't happen
08:59-!-Madmax [~oftc-webi@] has quit []
09:00*dzho looks at hyperbole-and-a-half
09:00<rsdehart>bug report all the things?
09:02<dubidub>Does /64 mean 2^64 number of IPv6 addresses?
09:05<nate>dzho: I'm not entirely sure, I would assume by now probably. It's apparently something to do with LE's backend ever since it enabled HTTP/2. HTTP 1.1 requested in cURL w/out --head freak out and crawl or timeout cause of content-length buggery
09:06<nate>So probably fixed in recent ones. This is CentOS so it's an old-ish cURL
09:07<nate>Meanwhile I wish there was a way you could issue an RSA + ECC at the same time
09:11*dzho browses through
09:13<nate>dzho: is the official fix in acme for it. There's several topics on the LE forums since around last September that end up pointing to this
09:13<dzho>nate: thanks!
09:14<nate>I think I first started seeing the issues around the end of 2018 though, at least pretty sure that's when I first started griping in here about it lol
09:14<nate>(not just with LE but various cURL using things)
09:15<nate>Guess the other things were all using HTTP/2 also. I'll have to try and see if I can fix cURL on here, oddly doesn't seem to be a package update
09:19-!-eyepulp [] has joined #linode
09:19-!-eyepulp is "eyepulp" on #linode
09:20-!-anomie [] has joined #linode
09:20-!-anomie is "Anomie" on #linode
09:34<Zr40>dubidub: it's actually 2^(128-64)
09:35<Zr40>dubidub: so a /56 contains 256 /64
09:38<Zr40>a /64 is special in that it represents a "network". What a "network" means exactly is up to you. It could be a vlan. It could be clients of a VPN server. It could be the trusted devices on your home LAN (so that they are separate from all the wifi gadgets that may or may not be secure). It could be something else entirely that you haven't thought of yet.
09:42<FluffyFoxeh>Why couldn't they have made it smaller?
09:42<Zr40>why should they?
09:42<Zr40>you have 2^64 _networks_. And a network contains 2^64 addresses
09:42<Zr40>both seem large enough to me that exhaustion isn't a problem
09:42<FluffyFoxeh>So that they don't waste 2^64 addresses on every network
09:43<Zr40>2^64 addresses in a network lets you do nice things, like stateless autoconfiguration
09:43<FluffyFoxeh>They could have made it so you could do SLAAC with fewer
09:44<Zr40>you assume it's a problem that needs solving
09:45<Zr40>remember that IPv4 has 2^32 addresses total
09:45<Zr40>you can fit the entire IPv4 address space in a single IPv6 /64... 2^32 times
09:45<rsdehart>this will be regarded as support of the claim that they should have been made smaller
09:46<Zr40>you don't save anything by reducing the address space
09:46<FluffyFoxeh>How many /64s are there?
09:46<FluffyFoxeh>Since they are allocated like IPv4 addresses were
09:46<rsdehart>2^32 times as many as there are IPv4 addresses
09:50<Zr40>even if you incorrectly assume that a /64 == one host (like it is with single IPv4 addresses) you still have 2^32 times as many hosts as there are v4 addresses
09:54<FluffyFoxeh>Single IPv4 addresses often represent multiple hosts via NAT
09:54<Zr40>not to the outside world
09:54<Zr40>there's only one port 80
09:56<Zr40>and I hope you're not putting your servers behind NAT
09:59<FluffyFoxeh>My home server is :p
09:59-!-memxor [] has joined #linode
09:59-!-memxor is "Kumar Kartikeya Dwivedi" on #linode #kernelnewbies #debian-next #debian
09:59<FluffyFoxeh>But this is a digression because it's not a problem with IPv6
10:00<FluffyFoxeh>And it still wouldn't be even if /112 was the standard allocation
10:00<FluffyFoxeh>or maybe go wild and make it /96
10:00<FluffyFoxeh>everyone gets a whole IPv4 address space :D
10:01<rsdehart>what would changing the allocation avoid?
10:02<rsdehart>I'm still trying to understand the nature of the perceived problem
10:02<rsdehart>Zr40 might be too
10:03<FluffyFoxeh>There might not be a real problem, it just "feels" wrong to have a 128 bit address space of which only 64 bits is effectively usable
10:03<rsdehart>s/might not be/is not/
10:04-!-kkd [] has quit [Ping timeout: 480 seconds]
10:04<chesty>it can also trick some people to believe /64 is twice as many addresses as /32
10:05<FluffyFoxeh>rsdehart: So there's no problem with /64, but on the other hand, would anything be lost if they had defined it to be /96?
10:05<FluffyFoxeh>They didn't, so this is hypothetical
10:06<rsdehart>yeah, 2^32 addresses per network
10:06<FluffyFoxeh>Are they not at least equally good choices?
10:06<rsdehart>and for what?
10:10<rsdehart>I'm just failing to envision the scenario that would lead to exhaustion of 18 sextillion allocations
10:11<chesty>you couldn't allocate an ip to every atom, for example
10:14<rsdehart>you couldn't even if a network was one ip
10:15<chesty>mind you, when they came out with ipv4 they probably couldn't envisage ever running out
10:15<rsdehart>right, but 18 sextillion vs 78 octillion
10:15<chesty>I'm not sure if ipv1 was /32, it probably was, so mind you, when they came out with ip14 they probably couldn't envisage ever running out
10:16<rsdehart>compared to ipv4 which maxes out at fewer addresses than there are people on the planet
10:24<linbot>New news from community: Does Linode have database products? <>
10:45<dubidub>Why does some customers need a /56?
10:47<Zr40>so you can have more than one /64
10:54<chesty>I'm not 100% sure, but I thought there was some ipv6 feature that required a /64, so if you wanted a few of them, they give you 256 of them
10:57<Zr40>that's the stateless autoconfiguration
10:58<Zr40>but you only need more than one if you need to have some hosts in a different /64
11:15<FluffyFoxeh>I wasn't arguing for changing it, just wondering about the rationale for making /64 the minimum for the cool features instead of something smaller
11:15<FluffyFoxeh>seems like it could have easily gone either way
11:15<FluffyFoxeh>there's a range of values that would work
11:15<FluffyFoxeh>they chose that one
11:19<FluffyFoxeh>they could have even made it larger!
11:33-!-Louis6321 [~Louis6321@2400:8901::f03c:91ff:fe37:abda] has quit [Ping timeout: 480 seconds]
11:33-!-Louis6321 [~Louis6321@2400:8901::f03c:91ff:fe37:abda] has joined #linode
11:33-!-Louis6321 is "Louis" on #linode
11:42-!-caubert [] has joined #linode
11:42-!-caubert is "Martin" on #linode
11:45<AlexMax>How often is the CentOS 7 image updated, and how is it built? (i.e. what determines what goes into the base installation and what is left off)
11:48<nate>Fairly certain all of the images are just the base images from the actual distros
11:48<nate>I don't think linode even does modified kernels anymore
11:49<nate>just some general pre-configuration for the network stuff and obviously root password
11:49<millisa> "created": "2014-07-08T14:07:21",
11:53<millisa>(used: curl -H 'X-Filter: { "label": {"+contains":"centos"}}' |python -mjson.tool )
11:53<@mcintosh>we still build our guest kernels
11:54<@mcintosh>they were never significantly modified though
11:54<@mcintosh>they've always been very close to the vanilla kernels
11:54<@mcintosh>created is... not what i'd key on for update frequency
11:54<millisa>it isn't updated when the image changes?
11:55<linbot>New news from community: Changing regions when deploying a bucket through the Linode CLI <>
11:55<millisa>(I have a feature request)
11:55<@mcintosh>created is not stamped, updated is
11:55<@mcintosh>for example, CentOS 7 was last updated on the 3rd of this month
11:56<millisa>don't see an updated in the /v4/images output
11:56<@mcintosh>it may not be there (perhaps it should be)
11:56<millisa>(feature request!)
11:56<@mcintosh>but it would be concerning if we hadn't updated the centos7 image at all since 2014 ;p
11:57<rsdehart>well, it is centos
11:57<rsdehart>it's just. that. stable.
11:58<@mcintosh>millisa: I opened a pull request to add that information (literally just now, because of you)
11:58<millisa>I blame AlexMax. He asked.
11:59-!-Peter [~oftc-webi@] has joined #linode
11:59-!-Peter is "OFTC WebIRC Client" on #linode
11:59<rsdehart>!point millisa
11:59<linbot>rsdehart: Point given to millisa. (130) (Biggest fan: mcintosh, total: 19)
11:59<rsdehart>!point mcintosh
11:59<linbot>rsdehart: Point given to mcintosh. (32)
11:59<AlexMax>Hopefully after you folks fix the API :P Having some issues
11:59<millisa>Peter: Greetings
11:59-!-Peter is now known as Guest18848
11:59<@mcintosh>AlexMax: "having some issues" is quite vague
12:00<@mcintosh>though someone internally just reported issues as well, so perhaps not just you
12:01<AlexMax>As in, getting 502 errors from API calls, admin panel just gives me three dots forever
12:01<Guest18848>How to resigned succes a account linode ??
12:01<AlexMax>in other words, broken in a way that I assumed you folks were already on top of, since "can't get into the linode manager" is probably a pretty high priority. :)
12:02-!-Guest18848 [~oftc-webi@] has quit [Remote host closed the connection]
12:12-!-ntox [~textual@] has joined #linode
12:12-!-ntox is "Textual User" on #linode #ovirt
12:15<linbot>New news from community: [Feature Request] Btrfs/XFS/... support <> || Changing regions when deploying an object storage bucket through the Linode CLI <>
12:16<@mcintosh>AlexMax: should be looking a little better now
12:25<linbot>New news from community: [Feature Request] Btrfs/XFS/ZFS/... support <>
12:46-!-sircmpwn [] has quit [Quit: Why do I even put this quit message in if I never quit]
12:50-!-dsapikas [] has joined #linode
12:50-!-dsapikas is "purple" on #linode
13:04-!-sircmpwn [znc@] has joined #linode
13:04-!-sircmpwn is "Drew DeVault" on #qemu #linode #ii #ceph
13:13-!-Shentino [~desktop@2600:6c54:7a00:133a:9a90:96ff:fece:3992] has quit [Ping timeout: 480 seconds]
13:14-!-sircmpwn [] has quit [Quit: Why do I even put this quit message in if I never quit]
13:15<AlexMax>volumes are still not available in Atlanta?
13:31<@pwoods>AlexMax: that's correct, BlockStorage is not available in Atlanta at the moment.
13:31-!-sircmpwn [znc@] has joined #linode
13:31-!-sircmpwn is "Drew DeVault" on #qemu #linode #ii #ceph
13:33<AlexMax>Hrm, it's been like that for as long as I can remember.
13:33<AlexMax>Just surprised that it hasn't made it over.
13:41<@pwoods>AlexMax: there's nothing to share at this moment, but there may be more on this in coming months.
14:54-!-ron_ [~oftc-webi@] has joined #linode
14:54-!-ron_ is "OFTC WebIRC Client" on #linode
14:54<ron_>Hello. How many domain can I add in a package? Is there any limitations?
14:58<Peng_>Linode provides VPSes, not shared hosting. There is no specific limit on domains.
15:01-!-andyzwieg103 [] has quit [Remote host closed the connection]
15:01-!-andyzwieg103 [] has joined #linode
15:01-!-andyzwieg103 is "azwieg103" on #linode
15:04<ron_>Thank you for your replay. Is operating a nanode plan is same as standard plan? Can I upgrade from nanode to standard plan?
15:14<ron_> Thank you again
15:15-!-ron_ [~oftc-webi@] has quit [Quit: Page closed]
15:17-!-Shentino [] has joined #linode
15:17-!-Shentino is "realname" on #tux3 #linode #mm #qemu
15:19<cruxeternus>!point Peng
15:19<linbot>cruxeternus: Point given to peng. (50)
15:40-!-doggycat [~u762141@] has joined #linode
15:40-!-doggycat is "n517245" on #linode
15:41<doggycat>i cant see a field to enter coupon codes at server checkout screen
15:46-!-doggycat [~u762141@] has quit [Remote host closed the connection]
15:50-!-doggycat [~u931732@] has joined #linode
15:50-!-doggycat is "n694458" on #linode
15:50-!-doggycat [~u931732@] has quit [Remote host closed the connection]
16:16<linbot>New news from community: Setup my domain to go to a node application with a different port? <>
16:20-!-andyzwieg103 [] has quit [Remote host closed the connection]
16:20-!-andyzwieg103 [] has joined #linode
16:20-!-andyzwieg103 is "azwieg103" on #linode
16:49-!-andyzwieg103 [] has quit [Remote host closed the connection]
16:54-!-aspis [] has joined #linode
16:54-!-aspis is "aspis" on #linode
17:00-!-anomie [] has quit [Quit: Leaving]
17:00<AlexMax>Does Linode have any solutions for hosting Docker images, or must I solve this on my own?
17:06<dzho>AlexMax: a good question!
17:06<dzho>I don't use Docker on linode, but I have found that many such questions are addressed, at least to some degree, in their online documents
17:27<@jdutton>AlexMax: You can absolutely utilize Docker on Linode. Here's a list of all of our containers guides, including multiple that talk about how to set up Docker:
17:55-!-fergtm [] has quit [Remote host closed the connection]
18:15-!-aspis [] has quit [Remote host closed the connection]
18:20-!-fergtm [] has joined #linode
18:20-!-fergtm is "Fernando" on #linode
18:29-!-aspis [] has joined #linode
18:29-!-aspis is "aspis" on #linode
18:31-!-Cromulent [sid301841@2001:67c:2f08:5::4:9b11] has quit [Remote host closed the connection]
18:31-!-tecuane [] has quit [Remote host closed the connection]
18:34-!-tecuane [] has joined #linode
18:34-!-tecuane is "Tecuane" on #linode
18:36-!-MrRobot7 [] has quit [Remote host closed the connection]
18:39-!-Cromulent [sid301841@2001:67c:2f08:5::4:9b11] has joined #linode
18:39-!-Cromulent is "Cromulent" on #linode #kernelnewbies
18:40-!-Cromulent [sid301841@2001:67c:2f08:5::4:9b11] has quit [Max SendQ exceeded]
18:40-!-Cromulent is "Cromulent" on #linode #kernelnewbies
18:40-!-Cromulent [] has joined #linode
18:40-!-MrRobot7 [] has joined #linode
18:40-!-MrRobot7 is "MrRobot7" on #linode
18:40-!-nate [] has quit [Quit: (?°?°)?? ???]
18:40-!-Cromulent [] has quit [Max SendQ exceeded]
18:40-!-Cromulent [] has joined #linode
18:40-!-Cromulent is "Cromulent" on #linode #kernelnewbies
18:45-!-eyepulp [] has quit [Remote host closed the connection]
18:46-!-jhitesma [sid129210@2001:67c:2f08:1::1:f8ba] has quit [Remote host closed the connection]
18:46-!-ericoc [] has quit [Remote host closed the connection]
18:48-!-jhitesma [] has joined #linode
18:48-!-jhitesma is "Jason Hitesman" on #linode
18:48-!-ericoc [] has joined #linode
18:48-!-ericoc is "ericoc" on #linode #dcs
18:50-!-theckman [sid295081@2001:67c:2f08:3::4:80a9] has quit [Remote host closed the connection]
18:50-!-zifnab [] has quit [Remote host closed the connection]
18:50-!-saintdev [] has quit [Remote host closed the connection]
18:51-!-zifnab [] has joined #linode
18:51-!-zifnab is "is wearing pants" on #linode
18:52-!-saintdev [] has joined #linode
18:52-!-saintdev is "Nathan Caldwell" on #linode
18:53-!-nate [] has joined #linode
18:53-!-nate is "Nathan" on #linode #php
18:54-!-theckman [] has joined #linode
18:54-!-theckman is "Tim Heckman" on #linode #dcs
18:56-!-purrdeta [sid201815@2001:67c:2f08:8::3:1457] has quit [Remote host closed the connection]
18:56-!-nyancat [sid3009@2001:67c:2f08:8::bc1] has quit [Remote host closed the connection]
18:56-!-mpontillo [] has quit [Remote host closed the connection]
18:57-!-nyancat [] has joined #linode
18:57-!-nyancat is "..." on #oftc #moocows #linode
18:58-!-mpontillo [] has joined #linode
18:58-!-mpontillo is "Mike Pontillo" on #virt #linode
18:59<@_brian>AlexMax I imagine Object Storage would be your best bet
19:00<@_brian>its also free until the end of the month if you wanna go try it out
19:00-!-purrdeta [] has joined #linode
19:00-!-purrdeta is "Alex" on #oftc #moocows #linode #help +#Corsair
19:28-!-Nebula_W2081d [] has quit [Read error: Connection reset by peer]
19:45-!-memxor [] has quit [Max SendQ exceeded]
19:45-!-kkd [] has joined #linode
19:45-!-kkd is "Kumar Kartikeya Dwivedi" on #linode #kernelnewbies #debian-next #debian
20:13-!-dsapikas [] has quit [Ping timeout: 480 seconds]
20:24-!-dsapikas [] has joined #linode
20:24-!-dsapikas is "purple" on #linode
20:31-!-aspis [] has quit [Read error: Connection reset by peer]
21:11-!-Winckle6 [] has joined #linode
21:11-!-Winckle [] has quit [Read error: Connection reset by peer]
21:11-!-Winckle6 is "The Lounge User" on #linode
21:11-!-Winckle6 is now known as Winckle
21:44-!-fstd_ [] has joined #linode
21:44-!-fstd_ is "fstd" on #gentoo #oftc #linode #debian #kernelnewbies
21:51-!-fstd [] has quit [Ping timeout: 480 seconds]
---Logclosed Thu Mar 12 00:00:33 2020