Sign In

Close
Forgot your password? No account yet?

Oh, look! Furaffinity is down...again! by Fluffball

Surprise, surprise. head desk

And, to make it all the more 'fun', there isn't a word from anyone in charge of the site (whether through their forums, or their Twitter accounts), giving the user base a head's up on what the issue may be or to give an ETA on how long it'll take to do whatever repairs the site needs this time.

Yay, transparency! rolls eyes

It's like they're trying to encourage people to up and leave the site.

Edit: From Furaffinity's Twitter...

'We recorded 56,000+ connections trying to hit FA to take us down. We went down briefly but we were able to make it though.'

Whoever made that note counted their chickens before they hatched, because the site is still down.

And kinda figured it was an attack (even Weasyl has been experiencing timeout errors; though not as bad as FA).

Seriously, I wonder if FA would be having their issues if they opted to use the protection package Cloudfare suggested for their site instead of the dirt cheap route that gave them limited protection against attacks.

Oh, look! Furaffinity is down...again!

Fluffball

Journal Information

Views:
323
Comments:
33
Favorites:
3
Rating:
General

Comments

  • Link

    Ah good, so at least it's not problems on my end. Maybe I should make updating my profile here a new year's resolution.

  • Link

    Yeah I'm not surprised in the least :/
    Neer made a comment on twitter but it's nothing special. Just saying they're 'looking into it'

  • Link

    It's starting to feel like they're testing to see how much their userbase will take before they'll finally stop drinking the empty promises kool-aid.

  • Link

    Re your update, yeah, I don't know how FA's currently using CloudFlare, but I was under the assumption based on their transparent and honest description of their donation drive and the use of the funds thereafter that they were going to use the funds to pay for the cost of CloudFlare. Instead it's going to ?????? because reasons.

    • Link

      DDoS attack mitigation is notoriously tricky and the site infrastructure must be designed for it. FA is definitely /not/ designed with DDoS resiliency in mind: there is only one central application server that handles user requests.

      Cloudflare right now isn't doing FA much good, since they can't cache replies to users' requests, so there's little traffic that can be prevented from reaching the central server.

      • Link

        FA's infrastructure has been held together with cellar tape and glue for years now. The fact that projects to replace it have been ongoing since 2006 is a testament to that. The issue isn't so much what CloudFlare will do for them as it is why it's still as broken as it is to the point that CloudFlare can't help them after all this time and after so many tens of thousands of dollars collected in donation drives. Let's not forget the fact that there is (was?) literally a server that was paid for in donation money sitting unused in Dragoneer's closet, and yet he's talking about buying more servers.

        One of the stated goals of the most recent donation drive was to hire professional coder(s) to actually fix stuff. Instead, it's being used to pay for "taxes" (read: Dragoneer's FA-accrued debt, the amount of which he stated to be $25k, the exact goal amount of the GoFundMe). Even Zidonuke wasn't "hired" - He wasn't being paid for his work. Neither will his replacement.

        • Link

          To be honest, I'm not following the FA news all that closely -- if what you say is true, then it's really cringe-worthy news.

          Personally, I've been monitoring the FA DevOps trello board; it sees quite a lot of (sensible) activity, so I'm moderately optimistic that the current batch of site admins at least seem to know how to keep things patched up.

          • Link

            There are some eye-openers there, too:
            https://trello.com/b/f7USczsS/furaffinity-devops

            • Link

              'Close all tickets from user who have been inactive for 3 months'

              ...I sincerely hope that doesn't mean they won't look into what was reported before they simply close the report to save themselves from doing extra work.

              • Link

                Eyeup. An eye-opener, right there, yeah.

      • Link

        Not to mention the fact that CloudFlare could be being used to route traffic so as to prevent FA's IP from being exposed to the public. While the current attack seems to be requests-based, the previous attack was not - It was directed at FA's actual IP, and then at their datacenter. While CloudFlare isn't foolproof, preventative measures could have been taken in both cases to minimize damage.

        • Link

          I don't claim FA is blameless! In fact, quite the reverse. :-)

          I'm just trying to point out that DDoS is something that even biggest Internet companies can have real hard time dealing with, not just FA. The interaction model of an art site only makes those problems harder to solve, too.

          That said, FA central web server is still it's critical weak spot; it can handle only so many connections before it chokes up, no matter whether there's an Internet-wide distributed filtering proxy in front of it or not.

          • Link

            It's why they should be investing in proxies placed across the globe to lessen the load on the single sever, but it seems like FA& Co. don't understand how servers are managed or even have a properly trained server team.

            • Link

              Proxies only help when they can cache content. In case of sites like FA, the page you are seeing is generated on the fly by the central server and only basic static content (like images) can be served from the cache. Except for small corner cases which are trivially optimizable anyway, each request must hit the central server anyway.

              There are ways to deal with that issue, but they require significant change in how the application backend is structured. I don't see that happening anytime soon, sadly.

              • Link

                There's still things that they can do with proxies and some cloud-based services that would alleviate almost all the traffic issues. Sure, it wouldn't stop the DDoS problems, but it would make it slightly more difficult as the spaced out servers can handle more traffic than just one. Also, regional centralized servers can be done, but they just need to make sure that they are talking with each other and the database servers to make sure all the servers know where everything is, but it's a low-level background action; it barely causes strain.

  • Link

    I'd comment to add to all of this, but OP and everyone else has basically said it all.. lol.

  • Link

    Hopefully this will get whomever is incharge of their servers to pull their head outta the ground and invest in more secure servers and the like.

    Won't be long before many of the FA goers come to Weasyl permanently. I imagine it also hurts you for your YCH and COmmissions? As well as other artists.

    • Link

      If they haven't done so now, they're not likely to do so any time soon - especially as long as Dragoneer is in charge.

  • Link

    I'm holding out hope more people will stop supporting a site that doesn't know how to support its users.

    • Link

      You'd think more people would leave - considering how much people have been donating to a site that promises so many changes in the last eight years, and have provided little to nothing in that regards. And when asked where the donation money went to, the staff is zip lipped (except in regards to the recent donations of almost 20k).

      • Link

        Yea, someone compiled a list of expenses that they noticed from Drag and Scig. It was not um.. not website-centric spending.

        • Link

          I only take that with a grain of salt. Drag and Scig's do have jobs; so whatever thing they buy, we can't automatically assume they're using donation money to pay for these frivolous things.

          However, I can see where the suspicions would come from; especially considering how long people have been donating to the improvement of the site, and we've barely seen anything done in that regards.

          • Link

            I think it's that history of 'where'd the money go' that makes people very, very skeptical. And reasonably so. If I donate to something, I like to see honest, transparent proof that my donation is going toward the thing I donated for!

            • Link

              Oh, I totally agree. That's why I never donate, as I never know exactly where the money is going.

  • Link

    It looks like FA is back up again. (I just logged in and did a few things.)

    • Link

      This journal was made yesterday morning, and the issues were resolved by then.

      • Link

        Ahh. Didn't even see it. But FA was down around 730 while I was on. Watched the guest count rise from 25k to 100k in less than 10minutes. I left before it went down. (if it did.)

        • Link

          It did go down, briefly (like, for half an hour?). All I could do was laugh- not out of amusement, but more out of amazement that they were still getting attacks (Weasyl, at the same time, seemed unaffected).

          From what I've heard, FA is in serious need of a complete overhaul/update so their DDOS can work properly.

          • Link

            It's because they've chosen the weakest version of protection. At least that's what I read. And it seems to be true since they're still falling to attacks.

  • Link

    FA seems to run on fairy farts and hobbit snot: collapses in a stiff wind. Somebody at the FA office opened a window, who was it?! XD

  • Link

    It seems to be down still for me. I don't know if its actually fixed yet or not.