HN is up again • Hacker News Board

Re: HN is up again

#341

Post by pmoriarty » Sat, Jul 09, 2022, 10:01 AM UTC

0xbadcafebee wrote: post #253

davedunkin wrote: post #66

Earlier quoted context omitted.

> Double disk failure is improbable but not impossible. It's not even improbable if the disks are the same kind purchased at the same time.

Even if they're not the same, they're written at the same time and rate, meaning they have the same wear over time, subject to the same power/heat issues, etc.

Hopefully, regularly checking the disks' S.M.A.R.T status will help you stay on top of issues caused by those factors.

Also, you shouldn't wait for disks to fail to replace them. HN's disks were used for 4.5 years, which is greater than the typical disk lifetime, in my experience. They should have replaced them sooner, one by one, in anticipation of failure. This would also allow them to stagger their disk purchases to avoid similar manufacturing dates.

Re: HN is up again

#342

Post by pmoriarty » Sat, Jul 09, 2022, 10:19 AM UTC

GekkePrutser wrote: post #230

grepfru_it wrote: post #174

Earlier quoted context omitted.

Slashdot has been around since 1997 and people still rave about its moderation system today. However, while I have high hopes for HN, it could very well go the way of digg overnight

I doubt that though. Digg was hyped way too much and the inevitable decline that comes after a hype killed it. Some things are good enough to survive that phase but Digg wasn't. HN never had a hype phase, just slow but strong & steady growth. And not growing too much either. It seems the perfect circumstances to really last. It doesn't have an invasive business model, or investors screaming for ROI either. That's the…

Change of ownership is inevitable, as people don't live forever. When that happens, if the new owners aren't interested or motivated to keep funding HN, it could easily go the way of the dodo.

Hopefully archive.org is involved in archiving HN, though unfortunately archive.org's future itself is in jeopardy.

Re: HN is up again

#343

Post by sschueller » Sat, Jul 09, 2022, 10:51 AM UTC

sillysaurusx wrote: post #20

HN was down because the failover server also failed: https://twitter.com/HNStatus/status/1545409429113229312 Double disk failure is improbable but not impossible. The most impressive thing is that there seems to be no dataloss, almost whatsoever. Whatever the backup system is, it seems rock solid.

Depends on your vendor as well.

A long time ago we had a Dell server which was pre setup raid from Dell (don't ask, I didn't order it). Eventually one disk on this server died, what sucked was that the second disk in the raid array also failed only a few minutes later. We had to restore from backup which sucked but to our surprise when we opened the Dell server the two disks had sequential serial numbers. They came from the same batch at the same time. Not a good thing to do when you sell people pre configured raid systems at a mark up...

Re: HN is up again

#344

Post by GekkePrutser » Sat, Jul 09, 2022, 12:06 PM UTC

pmoriarty wrote: post #342

GekkePrutser wrote: post #230

Earlier quoted context omitted.

I doubt that though. Digg was hyped way too much and the inevitable decline that comes after a hype killed it. Some things are good enough to survive that phase but Digg wasn't. HN never had a hype phase, just slow but strong & steady growth. And not growing too much either. It seems the perfect circumstances to really last. It doesn't have an invasive business model, or investors screaming for ROI either. That's the…

Change of ownership is inevitable, as people don't live forever. When that happens, if the new owners aren't interested or motivated to keep funding HN, it could easily go the way of the dodo. Hopefully archive.org is involved in archiving HN, though unfortunately archive.org's future itself is in jeopardy.

> though unfortunately archive.org's future itself is in jeopardy.

How so?? This is the first I've heard of it.

Re: HN is up again

#345

Post by justsomehnguy » Sat, Jul 09, 2022, 12:23 PM UTC

pmoriarty wrote: post #341

0xbadcafebee wrote: post #253

Earlier quoted context omitted.

Even if they're not the same, they're written at the same time and rate, meaning they have the same wear over time, subject to the same power/heat issues, etc.

Hopefully, regularly checking the disks' S.M.A.R.T status will help you stay on top of issues caused by those factors. Also, you shouldn't wait for disks to fail to replace them. HN's disks were used for 4.5 years, which is greater than the typical disk lifetime, in my experience. They should have replaced them sooner, one by one, in anticipation of failure. This would also allow them to stagger their disk purchases…

https://news.ycombinator.com/reply?id=32033520&goto=item%3Fi...

I've seen too many dead disks with a perfect SMART. When the numbers go down (or up) and triggers are fired then you are surely need to replace the disk[0], but SMART without warnings just means nothing.

[0] my desktop run for years entirely on the disks removed from the client PCs after a failure. Some of them had a pretty bad SMART, on a couple I needed to move the starting point of the partition a couple GBs further from the sector 0 (otherwise they would stall pretty soon), but overall they worked fine - but I never used them as a reliable storage and I knew I can lose them anytime.

Of course I don't use repurposed drives in the servers.

PS and when I tried to post it I received " We're having some trouble serving your request. Sorry! " Sheesh.

Re: HN is up again

#346

Post by pmoriarty » Sat, Jul 09, 2022, 12:37 PM UTC

GekkePrutser wrote: post #344

pmoriarty wrote: post #342

Earlier quoted context omitted.

Change of ownership is inevitable, as people don't live forever. When that happens, if the new owners aren't interested or motivated to keep funding HN, it could easily go the way of the dodo. Hopefully archive.org is involved in archiving HN, though unfortunately archive.org's future itself is in jeopardy.

> though unfortunately archive.org's future itself is in jeopardy. How so?? This is the first I've heard of it.

It's truly tragic how few people have heard of this. We should be doing all we can to raise awareness of this lawsuit before it's too late.

Here are some relevant links:

https://news.ycombinator.com/item?id=31703394

https://decrypt.co/31906/activists-rally-save-internet-archi...

https://www.courtlistener.com/docket/17211300/hachette-book-...

Re: HN is up again

#347

Post by jiggawatts » Sat, Jul 09, 2022, 1:06 PM UTC

bbkane wrote: post #296

jiggawatts wrote: post #293

Earlier quoted context omitted.

Is this... a deliberate attempt at constructing a Rube Goldberg machine? In all seriousness, at least 2/3rds of the complexity is because of your choice of tools and approach. Terraform alone makes things significantly more complex. If you just want to trigger a deployment, then a Template Spec made from a Bicep file could be banged out in like... an hour.[1] When in Rome, do as the Romans do. You basically took a Mi…

jiggawatts, this is an honest attempt. I'd LOVE it if there's an easier way I somehow missed. And talk is cheap. I dare you to write a blog post or make a public GitHub repo doing the equivalent work (see Goals section) with your own tools. If you can, I'll be super impressed (not that my admiration is worth anything ). One thing you'll run into is that AD roles and other authn aren't accessible via ARM templates/Bic…

> AD roles and other authn aren't accessible via ARM templates/Bicep

I normally bill for cloud automation advice, but the gist is:

You can automate RBAC/IAM via Bicep or ARM[1], but only for existing groups or system managed identities or user managed identities. This usually covers everything that is typically done for cloud automation.

Note that the initial setup might require "manual" steps to set up the groups and their memberships, but then the rest can be automated. In other words, there's a one-time "prerequisites" step followed by 'n' fully automated deployments.

You can also use templates to deploy groups dynamically[2] if you really need to, but this ought to be rare. The problem with this is that templates are designed to deploy resources, and AAD groups aren't resources.

More generally, your mistake IMHO was to try to automate the automation itself, while side-stepping the Azure-native automation tooling by choosing Terraform+Functions instead of Template Specs with delegated permissions via Azure RBAC. Most of your template is used to deploy the infrastructure to deploy a relatively simple template!

This reminds me of people writing VB Scripts to generate CMD files that generate VB Scripts to trigger more scripts in turn. I wish I was kidding, but a huge enterprise did this seven levels deep for a critical systems-management processes. It broke, and caused massive problems. Don't do this, just KISS and remember https://xkcd.com/1205/

[1] via Microsoft.Authorization/roleAssignments

[2] via Microsoft.Resources/deploymentScripts

Re: HN is up again

#348

Post by GekkePrutser » Sat, Jul 09, 2022, 2:05 PM UTC

pmoriarty wrote: post #346

GekkePrutser wrote: post #344

Earlier quoted context omitted.

> though unfortunately archive.org's future itself is in jeopardy. How so?? This is the first I've heard of it.

It's truly tragic how few people have heard of this. We should be doing all we can to raise awareness of this lawsuit before it's too late. Here are some relevant links: https://news.ycombinator.com/item?id=31703394 https://decrypt.co/31906/activists-rally-save-internet-archi... https://www.courtlistener.com/docket/17211300/hachette-book-...

Thanks! I heard of this back in the day but I didn't know it reached the courts now. I feel this is a bit self-inflicted though. A valuable service like this is already under scrutiny for copyrights and playing chicken with big publishers with tons of money to spend on lawyers is a really bad idea. I'm also opposed to the way copyright works but I would separate that fight from the service.

I guess it got them some goodwill during Corona but it could cause more damage than it's worth.

I wouldn't have done it, it was not like it was a real value during the pandemic. Those who are really into books and don't care about copyright already know their way to more gray-area sites like LibGen.

Re: HN is up again

#349

Post by scott_s » Sat, Jul 09, 2022, 2:18 PM UTC

dang wrote: post #303

sillysaurusx wrote: post #302

Earlier quoted context omitted.

Curiosity got the better of me. Why was there a 6 ID gap between the last post and first post? The answer seems to be that admins were making posts, which is neat. (There was also one lonely Flexport job ad.) Is your backup system tied to your API? Algolia is a third party service, and streaming the latest HN data to Algolia seems pretty similar to streaming it to a backup system.

I posted a bunch of test things and then deleted them.

I love this answer so much.

Re: HN is up again

#350

Post by coryrc » Sat, Jul 09, 2022, 3:45 PM UTC

lambda_dn wrote: post #266

olingern wrote: post #84

While the naysayers will say, "Why isn't this in the cloud?," I think the response times and uptime of hackernews is really impressive. If anyone has a write-up of the infrastructure that runs HN, I would be interested. Maybe startups really can be run off of a rasberry pi

HN was down for hours, no website hosted properly using cloud providers is down for more than a few minutes a year. It's trivial to set up multiple providers, multiple regions. Rather than having a few servers with some admin guy swapping out disks, really embarrassing for a so called tech site.

Hope you get a refund