(PREVIEW.GOATPEN.CO) OUTAGE NOTICE | 9/11/21 | ESTIMATED TIME >8 hours

[MadWorld]

Goatpen has a new account that is into foot fetish.

[shewhomustbeobeyed]

Goatpen has a new account that is into foot fetish.

EwwwwwwwwwwwWTFwwwwwwwwww!
Is this guy going to be the new MaroonSaint?
"Show me your feet." Ugh

[Serendipity]

Update on preview.goatpen.co

https://web.archive.org/web/20210912025 ... /post/6115

There have been some technical difficulties. We'll likely have 8 more hours (tops) of downtime tomorrow.

DriveItIn mod to main •

@Serendipity mod admin •

TL;DR Fuck computers and fuck their satan spawn horse shit. We’ll have at worst flat 8 hours down tomorrow starting around 10-12pm while I relocate the equipment.

OK.

With modern servers you run a special operating system called a hypervisor. In our case it’s ESXi. This operating system manages the resource allocation (ram, network, CPU) and acts as a broker\sheriff between multiple other machines(guests) that it manages. When you have more than one you want a way to unify management so you don’t have to go to multiple places to control the guests or the hosts. The software suite that manages this for us is vmWare.

Last night around ~2:30-3:00 I got a notification saying that the front end for the management system. vmWare vSphere was offline. I got home and sure as shit it was.

What I’ve come to realize over the last ~30-40 minutes is I probably fucked it up irreparably during troubleshooting when the problem was that it had lost connection to an NTP server and that time skew was what was preventing the services from coming up. I reregistered some service endpoints and recreated some certificates to do so. When the time skew was fixed services broke hard and I’m already at the end of my rope.

I delayed moving out by a day, basically just grinding the last 36 hours in between moving and cleaning and getting rid of all kinds of shit, and I have absolutely no choice but to at least tear down the old management system and bring it to two standalone hosts. This is probably going to take me another 4 hours.

I have to do this, and why I stopped moving my whole life is because in the sites design I leveraged something called a distributed switchport in order to have a software defined link aggregation group to go between the servers and the storage. This substantially increases performance. The only problem is that without vCenter server to tell the individual servers how to make their network switch function (at boot) the systems won’t know how to turn on the other computers. So this means if I power off the server I will have to do a substantial amount of work to even get to where I will be before I go to bed.

Anyway. My whole room is empty. I still have to wash the floors. This was me all day today.

Edit: here is a tardigrade porn that you can watch in the mean time

[Wowbagger]

But the guys running the site couldn't figure out how to disable the word filter?

For the record, the issue was that none of us had the time to learn Docker and migrate the database once we'd changed the code. I've since learned docker and made Lemmy my bitch but it's a piece of shit and I've left it in the dust. Quite recently, we've dusted off the idea of another Voat clone but we're in no hurry as frankly, shit posting is just as easily done in real time. You're welcome to join us anytime on our Matrix channels.

[Serendipity]

Update on preview.goatpen.co

https://web.archive.org/web/20210912025 ... /post/6115

There have been some technical difficulties. We'll likely have 8 more hours (tops) of downtime tomorrow.

DriveItIn mod to main •

@Serendipity mod admin •

TL;DR Fuck computers and fuck their satan spawn horse shit. We’ll have at worst flat 8 hours down tomorrow starting around 10-12pm while I relocate the equipment.

OK.

With modern servers you run a special operating system called a hypervisor. In our case it’s ESXi. This operating system manages the resource allocation (ram, network, CPU) and acts as a broker\sheriff between multiple other machines(guests) that it manages. When you have more than one you want a way to unify management so you don’t have to go to multiple places to control the guests or the hosts. The software suite that manages this for us is vmWare.

Last night around ~2:30-3:00 I got a notification saying that the front end for the management system. vmWare vSphere was offline. I got home and sure as shit it was.

What I’ve come to realize over the last ~30-40 minutes is I probably fucked it up irreparably during troubleshooting when the problem was that it had lost connection to an NTP server and that time skew was what was preventing the services from coming up. I reregistered some service endpoints and recreated some certificates to do so. When the time skew was fixed services broke hard and I’m already at the end of my rope.

I delayed moving out by a day, basically just grinding the last 36 hours in between moving and cleaning and getting rid of all kinds of shit, and I have absolutely no choice but to at least tear down the old management system and bring it to two standalone hosts. This is probably going to take me another 4 hours.

I have to do this, and why I stopped moving my whole life is because in the sites design I leveraged something called a distributed switchport in order to have a software defined link aggregation group to go between the servers and the storage. This substantially increases performance. The only problem is that without vCenter server to tell the individual servers how to make their network switch function (at boot) the systems won’t know how to turn on the other computers. So this means if I power off the server I will have to do a substantial amount of work to even get to where I will be before I go to bed.

Anyway. My whole room is empty. I still have to wash the floors. This was me all day today.

Edit: here is a tardigrade porn that you can watch in the mean time

Never saw this till now

I kinda wish I knew more about running a home server so I could point and laugh at this absolute trainwreck, I mean - all that for a site so shitty pix wouldn't even display without being cut off, AND a cuss filter?? I never did join that shit because my device wouldn't even allow me to register. Someone else made a name for me and it STILL didn't work.

Maybe that loss was for the best.

phpBB [video]

[MadWorld]

all that for a site so shitty pix wouldn't even display without being cut off, AND a cuss filter??

To be fair, lemmy can be very slow . Couple that with home server, you have a bandwidth problem. Lemmy hosts large snapshots and avatars of images locally. The text alone is about 300KB. The snapshot is roughly 200KB a piece and 20 pieces per page (worst case). The avatar is roughly the same, but appeared to be resized on the client side, which can be very slow. It would probably max out a 1gb connection under 100 users on a bad day.

Oh, and lemmy has community avatars, too, in addition to user avatars. Go figure

[antiliberalsociety]

What in the actual fuck??

And EVERYONE was shilling the shit out of Lemmy lol. I saw literally no good points to it.

[MadWorld]

What in the actual fuck??

And EVERYONE was shilling the shit out of Lemmy lol. I saw literally no good points to it.

I just checked some of the submissions on lemmy's default page. There were user avatars with dimension in the range of 1K x 1K. This one looked particularly bigger than the others.

https://lemmy.perthchat.org/pictrs/imag ... ee722.jpeg
"495508cc-f114-4795-a6c6-48286b2ee722.jpeg (JPEG Image, 3841 × 2161 pixels)"

And then it got resized down to the window of avatar, which is 32x32, on the client side!!

Each page has at least 20 user avatars and up to 20 community avatars. It does the same thing to the thumbnails.

That is a fuck load of big ass images to load!!!

And just as I thought I was done with it, I found another big ass thumbnail!!

https://lemmy.ml/pictrs/image/999344a9- ... 34c1d.jpeg
Transferred 2.68 MB (2.68 MB size)

What the actual fuck??? It burnt 2.68MB to load that big ass thumbview!!

Fucking mind-blown!! I almost shit myself

[antiliberalsociety]

Now imagine having to load each and every one of them, every time the chat page refreshes

[MadWorld]

Now imagine having to load each and every one of them, every time the chat page refreshes

Shit, it does the same thing to the comment section. This must be their unified model for loading all images

Ok, I misread your comment. Were you talking about talkloli's chatroom?