We are having serious problems with our auth and our database servers and this is causing cascading problems with mesh downloads and the forums. Sorry. We are working on this now and hope that it will be fixed in the next 12 to 18 hours.
Doesn't seem too bad at the moment, just a bit slow and periodically stops responding for a short while. Good luck to whomever is troubleshooting your DB server.
Issues can include the Tilemap is null!, No data for xx...etm, and various issues with flightor. Bare with us as it is worked on.
This shit happens before? Because... Just read: http://www.thebuddyforum.com/honorbuddy-forum/support-issues/75713-no-data-errors.html Date: 10.10.2012, 17:28
thank you for the update. though it did take you 12 hours Here's a crazy idea: If your company is selling a product that completely depends on 24/7/365 access to servers, then you to have a more serious PLAN for troubleshooting them, because servers of any sort will always, ALWAYS, get into trouble and require emergency reboots/fixes, and in the botting business they also tend to get ddos'd by hacker kiddies. Even the smallest, one-man internet companies will have some plan that pages them the moment their server goes down, or, they outsource this to a professional company. Instead here we get "One of our servers is going crazy. Waiting for Apoc to re-appear and have it either fixed or taken offline." seriously, your servers can just stop working for 12 hours and the ONLY solution you have is to wait for one person to hopefully become aware? what if he's on a 3-day weekend vacation. Last time the servers were "having issues" it took an entire week of "blacklisting node" crap till HB got back to normal. hopefully this will end up better, since obviously you aren't going to even consider letting people bot a Warsong Gulch or farm a herb without a constant connection to the (constantly failing) auth servers.
What could HB people do to make the bot more viable? Since these last couple of weeks it's been shitty due to authentication and meshes.
What can they do really though, system admins who can deal with custom high concurrency applications, web clusters (In this case PHP/MySQL (Clustered?)), file servers, and lord knows what else (distributed file systems? reverse proxies?) that's required to keep their setup up and running are not cheap to keep on the payroll, let alone kept awaiting possible 24/7 call-out. Edit: Not intentionally quoting just you, more the thread RE having to wait for Apoc to be called out and investigate.
I'm a conspiracy theorist, so I blame the nerd-ragers and fanbois. I think they're trying to hack the HB servers to take them over and mold them into what, and how, they [nerd-ragers and fanbois] think HB should run and work. Just my own little conspiracy theory, nothing more.
I secretly implanted a microchip into your brain to make you think like that. ... also bump this thread
I pressume you are using remote auth / mesh servers to avoid having people 'hack' your program. Since you cannot support this 'method' its best if yu find a more 'viable' authentication solution. the bot is really on alpha stages for more than 6 weeks now, I recall you saying the bot will be 100% functional in 1 week ( before mop hits ). I really have trust in you guys but opponents using other type of authentication ( eg shadow bot ) have 0 issues atm. Perhaps copy their authentication system?
Gb will not load old profiles is that part of it keeps telling me object not set to a instances of a refereance? says loading profile and sits there in uldum.
So, sometime around 11AM UTC our main auth gateway decided to stop responding. For whatever reason, our alerts never went off to let me know it happened. I've restarted that one, and taken down the current trouble-maker. At this point, I'm not sure what the cause was (there's absolutely nothing in our logs that points to any sort of issue with our gateway server), but I'll continue to monitor it. With the start of MoP (and a few of our other bots at the same time), the total traffic to our auth servers has almost doubled. Our servers are handling 10k+ requests per second on each of our main auth servers. Unfortunately, I think we're starting to hit the limits on available throughput. We've been working on a new auth system in the last few weeks that should hopefully be able to deal with the current traffic, and any future traffic increases, with far less resource usage (CPU/Memory/Bandwidth/etc). Unfortunately, that requires time, and testing. Bear with us folks, our poor old auth system is starting to show it's age and limits. We're working as best we can to get things sorted out. We're no happier about the issues than you guys are.
