TierHive General thread | Discuss, Updates, questions | LATEST: 8GB Ram available in some locations

Michal212 · June 6

The website does not state how much the paid static page program costs.

AnthonySmith · June 6

@Michal212 said:
The website does not state how much the paid static page program costs.

Damn that will teach me to tinker....

It's 0.000040 per 100MB (or part there of) per hour so basically about $0.35 per year.

And that's not replicated data for clarity if you have 100MB in 1 location or 100MB replicated in 10 locations we count that as 100 MB

Michal212 · June 6

What is the default character encoding forced by the hosting/server

stupidgenius · June 6

@AnthonySmith The registration flow seems like it has a bug or something. (No VPN)

I went to create an account and when I clicked "Create Account" nothing appeared to happen, I waited a moment, thinking it was an internet issue, but nothing changed so I clicked "Create Account" again, then the form cleared and I got " Verification failed. Please try again. " and I thought something went wrong with the Cloudflare captcha, so I tried to create an account again, and got the same error. So I thought maybe this is email verification, and sure enough I had an email sitting in my inbox, but that wasn't clear from the flow/error.

Finally, when I clicked the verification link the page took a strangely long time to load, but eventually said verification successful and I got logged in. Happy to provide more info if needed.

AnthonySmith · June 6

@stupidgenius said:
@AnthonySmith The registration flow seems like it has a bug or something. (No VPN)

I went to create an account and when I clicked "Create Account" nothing appeared to happen, I waited a moment, thinking it was an internet issue, but nothing changed so I clicked "Create Account" again, then the form cleared and I got " Verification failed. Please try again. " and I thought something went wrong with the Cloudflare captcha, so I tried to create an account again, and got the same error. So I thought maybe this is email verification, and sure enough I had an email sitting in my inbox, but that wasn't clear from the flow/error.

Finally, when I clicked the verification link the page took a strangely long time to load, but eventually said verification successful and I got logged in. Happy to provide more info if needed.

Thanks I will have a look through the logs, although we average around 50 registrations per day right now, hopefully everything is fine I would be surprised if no one else reported it but I will find your registration and check what happened

Fritz · June 7

Is there any limitation for free static hosting except storage?

Michal212 · June 7

@Fritz said:
Is there any limitation for free static hosting except storage?

Storage - 25 MB
Domains - 2
Zones - 3
Files-1,000

AnthonySmith · June 7

@stupidgenius said:
@AnthonySmith The registration flow seems like it has a bug or something. (No VPN)

I went to create an account and when I clicked "Create Account" nothing appeared to happen, I waited a moment, thinking it was an internet issue, but nothing changed so I clicked "Create Account" again, then the form cleared and I got " Verification failed. Please try again. " and I thought something went wrong with the Cloudflare captcha, so I tried to create an account again, and got the same error. So I thought maybe this is email verification, and sure enough I had an email sitting in my inbox, but that wasn't clear from the flow/error.

Finally, when I clicked the verification link the page took a strangely long time to load, but eventually said verification successful and I got logged in. Happy to provide more info if needed.

I looked through the logs and followed your registration through, 2 clicks were recorded on the registration within the same second, then again a 3rd click 21 seconds later.

The first 2 clicks were fast enough to work and actually attempt everything twice instead of blocking (due to inefficient code on our side)

The first 2 clicks actually triggered 2 Cloudflare Turnstyle attempts, but because our automation has already accepted you on the 1st click things were happening so you would not have got immediate feedback, and the last attempt is the last one you saw but because the first one already passed the 2nd/3rd failed all you saw the verification failed message from Cloudflare because it had already been validated on the 1st click.

Anyway, there is a gap of how long it takes from the first click to 2 API's responding and the reply being processed, so that if you click again, it will cause this result. Probably pretty rare for this to happen but we can fix it anyway with a few tweaks!

Thanks for the report. Bug report, thank you token applied, I appreciate it, its the little things that irritate people and we can fix it now we know

AnthonySmith · June 7

@Michal212 said:
What is the default character encoding forced by the hosting/server

We don't force encoding at the server level, as its a global service, files are served as is, you would need to declare it within the files, I guess you probably already know as you are asking the question but you would need to add the encoding into the head e.g. <meta charset="utf-8"> or within JS if needed.

stupidgenius · June 7

@AnthonySmith said: The first 2 clicks were fast enough to work and actually attempt everything twice instead of blocking (due to inefficient code on our side)

Sorry about the multiple clicks, didn't realized I had done that, but glad you figured it out. Thanks for the token!

rpqu · June 8

@AnthonySmith said:

@stupidgenius said:
@AnthonySmith The registration flow seems like it has a bug or something. (No VPN)

I went to create an account and when I clicked "Create Account" nothing appeared to happen, I waited a moment, thinking it was an internet issue, but nothing changed so I clicked "Create Account" again, then the form cleared and I got " Verification failed. Please try again. " and I thought something went wrong with the Cloudflare captcha, so I tried to create an account again, and got the same error. So I thought maybe this is email verification, and sure enough I had an email sitting in my inbox, but that wasn't clear from the flow/error.

Finally, when I clicked the verification link the page took a strangely long time to load, but eventually said verification successful and I got logged in. Happy to provide more info if needed.

I looked through the logs and followed your registration through, 2 clicks were recorded on the registration within the same second, then again a 3rd click 21 seconds later.

The first 2 clicks were fast enough to work and actually attempt everything twice instead of blocking (due to inefficient code on our side)

The first 2 clicks actually triggered 2 Cloudflare Turnstyle attempts, but because our automation has already accepted you on the 1st click things were happening so you would not have got immediate feedback, and the last attempt is the last one you saw but because the first one already passed the 2nd/3rd failed all you saw the verification failed message from Cloudflare because it had already been validated on the 1st click.

Anyway, there is a gap of how long it takes from the first click to 2 API's responding and the reply being processed, so that if you click again, it will cause this result. Probably pretty rare for this to happen but we can fix it anyway with a few tweaks!

Thanks for the report. Bug report, thank you token applied, I appreciate it, its the little things that irritate people and we can fix it now we know

There's so many checkboxes to click. How long until catch up to # of tierhive PoP 🤣🤣🤣🤣🤣?

budi1413 · June 8

singapore has fallen?

AnthonySmith · June 8

@budi1413 said:
singapore has fallen?

its back

Neat321 · June 8

Palermo, Italy location 👉👈

AnthonySmith · June 8

@Neat321 said:
Palermo, Italy location 👉👈

Can you suggest a host?

Neat321 · June 8

@AnthonySmith said:

@Neat321 said:
Palermo, Italy location 👉👈

Can you suggest a host?

Maybe melbicom or datapacket

budi1413 · June 9

hi ant, in the billing > token use history section, the date & time column timezone doesn't follow timezone setting at the account settings.

AnthonySmith · June 9

@budi1413 said:
hi ant, in the billing > token use history section, the date & time column timezone doesn't follow timezone setting at the account settings.

I will have a look.

budi1413 · June 10

@AnthonySmith said:

@budi1413 said:
hi ant, in the billing > token use history section, the date & time column timezone doesn't follow timezone setting at the account settings.

I will have a look.

also at the vps instance > performance graph. sorry for my ocdness.

skorous · June 10

@budi1413 said:

sorry for my ocdness.

Can't speak for Ant obviously but I'd appreciate the feedback since since it's minor enough most people probably won't bother/notice.

AnthonySmith · June 10

@skorous said:

@budi1413 said:

sorry for my ocdness.

Can't speak for Ant obviously but I'd appreciate the feedback since since it's minor enough most people probably won't bother/notice.

Yeah tell me everything 😀 the best bug reports right now are the "it's annoying when" or "it's confusing when" because the next big push is UI/UX

WSS · June 10

@AnthonySmith said: Yeah tell me everything

My balls itch, and they weren't before I had to login using IPv4 to update some packages.

AnthonySmith · June 10

@WSS said:

@AnthonySmith said: Yeah tell me everything

My balls itch, and they weren't before I had to login using IPv4 to update some packages.

Ipv6 is on 2.1% now it's not looking good.

John_Q_Developer · June 10

@WSS said:

@AnthonySmith said: Yeah tell me everything

My balls itch, and they weren't before I had to login using IPv4 to update some packages.

Switch to ipv6 only.

WSS · June 10

@John_Q_Developer said:

@WSS said:

@AnthonySmith said: Yeah tell me everything

My balls itch, and they weren't before I had to login using IPv4 to update some packages.

Switch to ipv6 only.

I'm still IPv4 only at work because I don't want to have to manage vending IPv6 to anyone here, little alone the guest network. Not even a punched-through hurricane /48 for myself.

AnthonySmith · June 10

@budi1413 should be ok now

Timezones for token use and performance metrics are not respected.

Also the recipes deployment flow should be a bit better now, somehow as v0.000000001 of that made it in to live because it just worked we never touched it again, but its confusing, so the UI update is in.

TemporiousOne · June 11

@AnthonySmith said:

Also the recipes deployment flow should be a bit better now, somehow as v0.000000001 of that made it in to live because it just worked we never touched it again, but its confusing, so the UI update is in.

The un-intuitive necessity to click on "Calculate Cost" before deployment of recipes is gone. Dynamically generated cost. Excellent!

xms · June 11

@AnthonySmith said: Ipv6 is on 2.1% now it's not looking good.

I am surprised it passed 1%.

Currently for first time users.

One need to get a IPv6 prefix. (one time per location)
Create VPS.
Allocate a single IPv6 address.
Attach it to the VPS.
Go find the IPv6 setup article which is buried DEEP, its not a linked in the VPS page. https://tierhive.com/blog/technical_articles/configuring-ipv6
Manually fill the blanks, for the config.
Login via IPv4 address in the console.
Run the ipv6 enable commands.
If anyone want to login via SSH via IPv6, its not in the UI. Guess again.

Vs for IPv4.

Create VPS.
Login via the ssh shown in the UI.

You make it extra hard because you dont like it, and it shows.

Just my advice: dont let your personal opinions affect your business. I made that mistake, it stalled the growth for several years.

AnthonySmith · June 11

@xms said:

@AnthonySmith said: Ipv6 is on 2.1% now it's not looking good.

I am surprised it passed 1%.

Currently for first time users.

One need to get a IPv6 prefix. (one time per location)

Create VPS.

Allocate a single IPv6 address.

Attach it to the VPS.

Go find the IPv6 setup article which is buried DEEP, its not a linked in the VPS page. https://tierhive.com/blog/technical_articles/configuring-ipv6

Manually fill the blanks, for the config.

Login via IPv4 address in the console.

Run the ipv6 enable commands.

If anyone want to login via SSH via IPv6, its not in the UI. Guess again.

Vs for IPv4.

Create VPS.

Login via the ssh shown in the UI.

You make it extra hard because you dont like it, and it shows.

Just my advice: dont let your personal opinions affect your business. I made that mistake, it stalled the growth for several years.

It's going to be automated as part of the VPS deployment, but that's not simple and we are a small team and ipv6 was never planned for alpha.

There are copy paste command examples now in the faq the blog is just extra info.

I don't hate ipv6 really, but I do think there is disproportionate noise made about it Vs actual use.

I admit adding ipv6 in TierHive is not frictionless but the reality is about 15 - 30 seconds of actual effort, 98% never even issue a /64 without knowing additional steps are needed beyond that. I read that as people don't actively want ipv6.

Anyway, longer term, it will be automatically added during deployment, we just need to pick what development time is spent on efficiently at the moment.

WSS · June 11

That chip on Ant's shoulder grows bigger every time someone mentions how he hates IPv6.

Poor bloke just doesn't want to enable Internet of Toasters.

Scioner · June 11

Not sure if it's worth mentioning, but it seems like recent komari-agent (monitoring for servers, like nezha) update somehow crushed my instance.

It took two reboots from management panel to recover from that. But it's working just fine since.

I'm not sure honestly what was the cause. It was a clear pattern - update>issue. There was no spikes in cpu or memory usage. And the same piece of software works after two reboots.

So I'm reporting it just in case.

WSS · June 15

Hey @AnthonySmith -

I've been throwing 444s instead of 403s for my main site, but you're probably getting a ton of HAProxy 502s because of this whenever it gets scanned, right? Want me to behave and turn it back to 403s?

AnthonySmith · June 15

@WSS said:
Hey @AnthonySmith -

I've been throwing 444s instead of 403s for my main site, but you're probably getting a ton of HAProxy 502s because of this whenever it gets scanned, right? Want me to behave and turn it back to 403s?

Don't think that matters to us, as long as your main domain responds with a 200 we pass the traffic through for you to deal with.

I will check with my colleague tomorrow he is the guru on this.

WSS · June 15

Well, if I throw a 444 and HAProxy is expecting some sort of response, I don't think it's going to be all that happy by me just closing the connection. A 403 is at least something it expects.

AnthonySmith · June 16

Haproxy update to hopefully iron out some false negatives.

We were checking your backend every 2 minutes for a 200,301 or 302 response if it failed 10 times in a row we removed it from the live haproxy config.

That has been changed to 60 in a row (2 hours)

We are also looking to add a user side override a "trust me bro" mode for none load balanced records

Encoders · June 17

is it just me or kernelkeeper site is pitch black dark

ms edge 149.0.4022.69 in kubuntu. works fine in firefox tho

AnthonySmith · June 17

@Encoders said: ms edge 149.0.4022.69 in kubuntu.

... thanks I will add that to the list although I have to say... MS Edge on kubuntu was not really ever a test case

somik · June 17

@AnthonySmith said:

@Encoders said: ms edge 149.0.4022.69 in kubuntu.

... thanks I will add that to the list although I have to say... MS Edge on kubuntu was not really ever a test case

MS Edge on Mac is a valid browser validation used by some top companies. But I don't remember it being used on linux... Firefox and Chrome on Debian and Ubuntu were tested...

WSS · June 17

I remember testing against OmniWeb on NeXT.. in the 2000s. It was a hell of a time.

AnthonySmith · June 17

@WSS said: NeXT

I just said "Oh My God" out loud involuntarily haha

WSS · June 17

@AnthonySmith said:

@WSS said: NeXT

I just said "Oh My God" out loud involuntarily haha

With 96MB of RAM, an 040/50Mhz overdrive processor designed for a Mac (barely fit), and a 4GB SCSI drive, I was living high in hog heaven. There were some minor timing issues with the CPU upgrade so it'd stutter for about a second every hour or two but was a hell of a lot faster than the 040/25 it came with.

I did end up getting the Y2K CD patchs from Apple for 3.3 and 4.

Never did have a BeBox. For some reason was just never driven (or rich enough).

AnthonySmith · June 17

ISO over https coming soon, hopefully tomorrow!

WSS · June 17

@AnthonySmith said: ISO over https coming soon, hopefully tomorrow!

MJJs are going to abuse love this.

cmeerw · June 17

@AnthonySmith said: ISO over https coming soon, hopefully tomorrow!

BTW for "press ESC at boot" - usually that requires a lot of luck and good timing to manage to connect to the VNC in time and be able to press ESC.

AnthonySmith · June 17

@cmeerw said:

@AnthonySmith said: ISO over https coming soon, hopefully tomorrow!

BTW for "press ESC at boot" - usually that requires a lot of luck and good timing to manage to connect to the VNC in time and be able to press ESC.

We have it set to 20 seconds instead of the default 3 hopefully it will be fine

Not_Oles · June 17

@AnthonySmith said:

ISO over https coming soon, hopefully tomorrow!

This is great! Thank you!

Now my 256 MB RAM TierHve VPS is expected to have sufficient extra space temporarily allocated from the node to mount an ISO!

Unless I am missing something. Thanks again!

AnthonySmith · June 17

@Not_Oles said:

@AnthonySmith said:

ISO over https coming soon, hopefully tomorrow!

This is great! Thank you!

Now my 256 MB RAM TierHve VPS is expected to have sufficient extra space temporarily allocated from the node to mount an ISO!

Unless I am missing something. Thanks again!

It mounts/streams the iso directly over https so there is no extra space given or needed.

Not_Oles · June 17

@AnthonySmith said:

It mounts/streams the iso directly over https so there is no extra space given or needed.

I imagined that additional RAM space is allocated from the node for the ISO while the ISO is mounted and being accessed. If I understand right, it's not like the RAM space for the ISO comes from and is limited by the RAM allocated by the node to the VPS.

Please correct me if I am wrong. Thanks again!

Tom

WSS · June 18

If you can't fit something into RAM, why do you think you would be able to over the internet?

somik · June 18

@AnthonySmith said:
ISO over https coming soon, hopefully tomorrow!

Enough time to find my windows XP ISO

TierHive General thread | Discuss, Updates, questions | LATEST: 8GB Ram available in some locations

Comments