V-3YG7 Fortizar Fight – Servers vs Everyone

Stephanie Daugherty 2017-04-17

On Friday, April 14, a battle over a final timer for a Fortizar brought out over 2000 pilots, and sent the Imperium home empty-handed as a server node plagued by soul-crushing lag and stuck modules allowed the structure to repair itself despite forces gathered to destroy it.

The Imperium had set out from Delve to attack a Circle-of-Two Fortizar in V-3YG7 in Catch, taking along just over 1000 pilots. They were met by a similar number of defenders, and with third parties joining in the fun too, the numbers in local hit 2100.

On the subcap side, TEST brought out a Nightmare Fleet…

Nightmares in flight

…whilst both The Imperium and CO2 brought Machariels, and Ferox fleets were present from Soviet Union and Red Alliance.

Ferox Close Up

Fax and Ferox in combat

Machariels Tethered on the citadel

Imperium FC Apple Pear told us that The Imperium were hitting the damage ceiling on the Fortizar, and then the defenders cynoed in carriers and launched fighters. This extra load, combined with the fighters being launched by The Imperium, caused the node to have issues. Pilots started reporting socket closures, and red-cycling of modules, which is a desync between the server and the client – rather than indicating the actual state, it appears to be cycling while flashing red. When this happens, It becomes impossible to toggle the module on or off, or even to see what state it is in. This condition happens in some fights when time dilation isn’t enough to stabilize the server node and communication between server and client eventually breaks down entirely.

It was noted on Reddit that Imperium carriers pulled their fighters off the citadel and onto other ships on the field at this point. Under the server conditions, only the citadel’s own defenses and carrier fighters were working somewhat reliably, but as the Fortizar’s point defense system could make quick work of any fighters within its reach, Imperium FCs decided not to waste their only working weapons.

The server problems led to some strange side effects, such as one of the Force Auxiliary ships belonging to the defenders being effectively unkillable due to its Capital Emergency Hull Energizer cycling for over an hour. Weapons continued to use ammunition, even without a target locked, and Logistics pilots on both sides reported serious problems when trying to rep their targets.
Pilots who were hit by bombing runs were also, due to the desyncs and red-cycling, able to warp off when they ordinarily would have been a wreck on the field. One TEST pilot reported being hit by a full barrage of bombs and still being able to warp off in armor.

Despite this, there were multiple successful bombing runs from both sides and the third parties. The Imperium’s DaBigRedBoat led several of these runs, including one which landed on an FCON Tempest fleet.

The server problems continued even after the fight had finished, with defenders who stuck around to pick up Imperium stragglers and pilots who had disconnected coming across an Archon who they managed to get into hull, the pilot activated his emergency energizer, and, the same thing happened as with the Force Auxiliary ship earlier in the fight, the module became stuck on, making the carrier invulnerable for 30 minutes, while the pilot was able to launch his Fighters at the attackers.

One other thing, at an hour into this grind, a lone Helios foolishly tried to evac blueprints from the citadel, only to be caught as they left.

Quite how the pilot had not been aware of the upcoming timer in the preceding six days, or the fact that a large fight was likely to kick off is unknown, but, it is a brave pilot who tries to evac blueprints during a 2100-pilot Tidi fest.

Just about the only person on field unaffected by the server problems was our photographer Razorien, who was able to take a lot of pictures of explosions and fleets mid-warp, some of which you can see in this article, and the rest can be found on his Flickr.

Let your voice be heard! Submit your own article to Imperium News here!

Would you like to join the Imperium News staff? Find out how!

Comments

  • Rhivre

    Thanks Kyalla 🙂 Was the node not reinforced, or was it just the sheer number of pilots + fighters/drones/explosions that were causing the problems?

    April 17, 2017 at 1:06 pm
    • Stephanie Daugherty Rhivre

      as I understand it, the node was not reinforced.

      April 17, 2017 at 3:50 pm
      • It was not, as system next door had identical TiDi which is a good indicator of a shared node.

        April 18, 2017 at 10:38 pm
  • Bill Bones

    CCP should just give up on this “bring pilots until the server breaks” silliness and stop trying to calculate these combats in “real time”. They should just add a turn-based combat system for large fights and call it a day.

    April 17, 2017 at 1:39 pm
    • Axhind Bill Bones

      There was some idea where you could do multi threading of the fights. Basically, take all commands in a tick, calculate damage/rep for each thing spread out among the cores and then just take the end result as good enough. That means that commands would not be analysed in the sequence they arrive at the server but total amount of damage vs reps during the tick decides how the ship/structure looks at the end of the tick. There are probably plenty of edge cases with this approach but could be something that helps in big fights.

      April 18, 2017 at 10:37 pm
      • Bill Bones Axhind

        99% of EVE’s performance issues stem from the same: it uses single-threaded code. That’s how it was coded back in 1999-2003 and updating it to multithreading would require rewritting everything from scratch, which would tantamount to develop “EVE 2”.

        April 19, 2017 at 1:35 pm
        • Axhind Bill Bones

          True and the system suggested above doesn’t really behave the way it does now where commands are executed in order they are received so timing does matter for ship survival even if all actions happened during the same tick (as far as I understand it).

          TiDi also took a long while for CCP to accept as only possible solution. Maybe this idea has merits too.

          April 19, 2017 at 9:03 pm
  • annon

    it wasn’t the final timer it was the armor timer and lets be fair the issues with the server will have been a problem for both sides of the fight not just the attackers.

    April 17, 2017 at 4:18 pm
  • Not A Fan

    Seriously?!? The Goons exploit this mechanic all of the time in order to defend their timers but when it works against them they want to cry about it?

    April 17, 2017 at 4:42 pm
    • Rhivre Not A Fan

      I dont see crying, I see both sides having problems with the desyncs. So, technically I guess both sides are crying.

      April 17, 2017 at 4:46 pm
    • SeaBassSA Not A Fan

      The only tears I see here are from your manufactured outrage.

      April 17, 2017 at 5:36 pm
    • dragonshardz Not A Fan

      “Goons exploit this mechanic all of the time in order to defend their timers”

      [citation needed]

      I have not seen a single Imperium citadel timer where we’ve taken advantage of Citadel repair timers desyncing with the rest of the server, save in the very earliest days of Citadels at the end of the Casino War.

      April 19, 2017 at 11:14 pm
  • Daito Endashi

    Wow, the screenshots are truly awesome

    April 17, 2017 at 6:54 pm
  • Tsyolin

    Yeah it was a pretty good fight minus all the tidi cancer involved, the problem for the goons & co was that when we made the decision to engage they pulled all of their damage off the fort. Some said it was poor communication, or pilots not listening to directions. Either way, it was way too late by the time the subcap fleet starting shooting the fort again.

    Another mistake was clearing the field of subcaps before getting the entire cap fleet out, some cloaky dictors managed to bubble like 7-8 carriers/FAX that died quietly. (Minus the little Minokawa that could, that was an experience haha).

    April 17, 2017 at 8:11 pm
    • Stephanie Daugherty Tsyolin

      Wasn’t there personally, but interviewed a number of the FCs and other people that were there for the article. As I understand it, the carriers pulled their fighters off because they couldn’t keep them from being hit with the point defense, and most subcaps were red cycling by this point.

      April 17, 2017 at 8:17 pm
      • Yeah carriers pulled their fighters back when the fort was at about 2 ticks left of Armor, most goons I spoke to said the subcap damage should have been more than enough but server problems combined with too many people switching targets screwed them over.

        April 17, 2017 at 8:45 pm
      • To be honest fighters worked so badly with that it’s questionable if their DPS would have helped. Tons of them red cycled and never got back. Strange thing is that damage output dropped like a rock in a black hole gravity well but you could still command them to move around and recall with perfect response by the server. Seems that in TiDi fights it’s all about who can recall/launch fighters do one volley, recall/launch faster. Very weird behaviour. It used to be that guns on fighters would work well due to being server side but now it’s almost the exact opposite.

        April 18, 2017 at 10:32 pm
      • We had orders to keep one wing of fighters on the Fortizar and keep the other two further back. The wings applying DPS were safe, but the rest received bomb damage.

        April 19, 2017 at 7:38 am