Merging Servers ~ The Full Facts

2»

Comments

  • Father_gold - Sanctuary
    Father_gold - Sanctuary Posts: 449 Arc User
    edited May 2015
    where there is a will there is a way.

    but if the game was set up during a time before things like merges or the current problems were really a thing... that is the consequence of an old game.

    i mean they would really have to be able to deal with the alt issue in order to do any good merging. it would be more productive to fix/develop/implement server- migration instead. if at a cost, there would less alt moving. but this option also has some of the same problems.
  • Boomz - Harshlands
    Boomz - Harshlands Posts: 516 Arc User
    edited May 2015
    sylen can you give us the full lay down similiar to this thread on the feasibility and as a result a scale of possibility from 1-10 on a server transfer stone? This would effectively solve the same issue and give pwi money in the process if said person wishing to transfer wanted to. I would assume the only major database conflictions would be names, no?
    ░░░░███████]▄▄▄▄▄▄▄▄
    ▂▄▅█████████▅▄▃▂ cause i can't make art, so i made
    ◥⊙▲⊙▲⊙▲⊙▲⊙▲⊙▲⊙▲⊙▲⊙◤this awesome tank.
    [SIGPIC][/SIGPIC] 1 shot king. LV 150 FSP bosses hit for 3m. Top Player hit record: 652,656.
  • NeonZephyr - Archosaur
    NeonZephyr - Archosaur Posts: 178 Arc User
    edited May 2015
    Excellent thread Sylen,

    Seems it would take way too much effort for a merge, wouldn't people rather that time, effort and money to be focused on fixing bugs, stabilizing the economy, giving us a better support team and just overall making the game enticing enough to bring the populations up naturally?

    Servers don't need a merge for population, I want a better game to play and I don't think everyones money should be wasted on technical acrobatics.
    it should be because it is.
  • SylenThunder - Twilight Temple
    edited May 2015
    sylen can you give us the full lay down similiar to this thread on the feasibility and as a result a scale of possibility from 1-10 on a server transfer stone? This would effectively solve the same issue and give pwi money in the process if said person wishing to transfer wanted to. I would assume the only major database conflictions would be names, no?

    Basically you would run into the same issues, and no, it's even more integrated than just the name database.

    In the proprietary databases you will have all the links in your friends list, faction data, relations, ect. If you remove your character data from there, and import it to the databases on a different server, you end up with your account no longer functional because it cannot tie into the old relational data any longer, and the database you removed it from will eventually corrupt as the other data attempts to access your old account info.

    It is exactly this type of thing that led me to the discoveries of how it all works, and to the information that I posted at the beginning of this thread.
    I had my private server, and I was attempting to transfer account data from an old server, to a newer one I created. It pretty much came down to either all, or nothing. The inter-relational ties between the databases were just too complex for me to separate the data without corrupting it and making it so that either the character was unplayable, the account was unplayable, or crashing the server entirely.

    When the two servers share the same databases, it would probably be easier, but still will be too difficult of a task to be automated reliably. I was able to do it successfully in most cases when they ran off the same databases, but it only had about a 90% success rate.
    [SIGPIC][/SIGPIC]
  • LucySong - Sanctuary
    LucySong - Sanctuary Posts: 9 Arc User
    edited May 2015
    From what I see even if merge Sanc, Arch. HT all together will not make big difference tho, Every1 plays at they on time, Not sayin that will not give extras ppl or w/e... But thinkin of what can go wrong is bigger than actually working... 5-6 years of probs no matter what our devs can or can't do. b:surrender ex: Lycaeum still broken and there's more...
  • trands
    trands Posts: 2
    edited May 2015
    This thread is so upside-down. The whole starting point is "should the population of server X and Y get brought together". None of the factors mentionned here exceed the point of minor importance. This is simply a show-off about Sylen being into programming stuff and how much he dislikes the big TW faction(s) on his server.

    If any of the servers that I don't play is at that state of needing a "merge", I do not know and do not care. However, this thread is pointles. The job of technicians should be to find options or alternatives, each having imperfections and relative costs, so they can be confronted with the potential gains. Stating "there is no perfect way" is silly. "School-case" knowledge has no value if it can't open it to the real world and confronted with reality.
  • Vanflyheight - Heavens Tear
    Vanflyheight - Heavens Tear Posts: 420 Arc User
    edited May 2015
    trands wrote: »
    This thread is so upside-down. The whole starting point is "should the population of server X and Y get brought together". None of the factors mentionned here exceed the point of minor importance. This is simply a show-off about Sylen being into programming stuff and how much he dislikes the big TW faction(s) on his server.

    If any of the servers that I don't play is at that state of needing a "merge", I do not know and do not care. However, this thread is pointles. The job of technicians should be to find options or alternatives, each having imperfections and relative costs, so they can be confronted with the potential gains. Stating "there is no perfect way" is silly. "School-case" knowledge has no value if it can't open it to the real world and confronted with reality.

    Would you like to see this thread closed due to your troll factor of stupidity? Plus this isnt something that should be trolled over, its only posted for information.
    The Sure Shot that Flies Straight

    Tiduswarrior Demon 101 (Main), Vanflyheight 100 (Demon RB2), SasukeZx 95 (Demon), Leobeastking 90s (Sage), Swiftterror 80s, AquaStriker 99 (Sage)

    2nd Acc: BlademageX 88, RazorFalcon 89, RavenwingZ 79, Veilpor 73, TidalLight 30, SythrilZ 64, Stormthril 64
  • ovenusarmanio
    ovenusarmanio Posts: 6,695 Community Moderator
    edited May 2015
    trands wrote: »
    This thread is so upside-down. The whole starting point is "should the population of server X and Y get brought together". None of the factors mentionned here exceed the point of minor importance. This is simply a show-off about Sylen being into programming stuff and how much he dislikes the big TW faction(s) on his server.

    If any of the servers that I don't play is at that state of needing a "merge", I do not know and do not care. However, this thread is pointles. The job of technicians should be to find options or alternatives, each having imperfections and relative costs, so they can be confronted with the potential gains. Stating "there is no perfect way" is silly. "School-case" knowledge has no value if it can't open it to the real world and confronted with reality.

    PWI is a business. Economic feasibility and profitability are something that businesses should always taken into consideration when making business decisions. Many businesses have crumbled because they took on more than their business could realistically handle. So, I'd say, this is absolutely good information to know.

    kritty%20sig_zpsp0y7ttsb.png
    Thanks to MikoTenshi for the Avi and Kritty for the Signature.
    Get the Forums Enhancement Extension!
  • heerohex#3018
    heerohex#3018 Posts: 4,795 Community Moderator
    edited May 2015
    In any case this is a possibility. Im sure a program could be made to deal with the problems at hand and a workaround found. Lets put it like this "I don't believe in no win scenarios" Some of you will know where that comes from, some not, but when theirs a will their is a way.

    Either way at this point yes its going to be hard work. You can all talk about costs as much as you want but bottom line I wonder how much PWI could save from doing this over Server Costs THE list here would be endless. ( Weighing up the Costs here )

    Server Costs INC Repair and Replace.
    Mant Costs for servers. ( Someone has to look after them )
    Bandwidth Costs.
    Any software Costs (INC Licence Per Server Or Upgrades)
    Player base.
    Server Location Costs.
    Power Costs.
    Sales from redundant equipment. (Or reuse)
    Merger Costs.
    Income From Merged Server
    ETC will all play a roll in this coming down to it.

    The point is clear this could be between shutting PWI Down and Keeping It going one of the many answers.

    I often reply in Suggestion Box with the same comments "I would like to say the GMs are watching this as well as everything on a good continuous basis. When the time comes I would like to trust that it will be done in the right way." for the players and Company.

    Really Really gotta take a lot more into account here...
  • unrefuted
    unrefuted Posts: 107 Arc User
    edited May 2015
    Basically you would run into the same issues, and no, it's even more integrated than just the name database.

    In the proprietary databases you will have all the links in your friends list, faction data, relations, ect. If you remove your character data from there, and import it to the databases on a different server, you end up with your account no longer functional because it cannot tie into the old relational data any longer, and the database you removed it from will eventually corrupt as the other data attempts to access your old account info.

    It is exactly this type of thing that led me to the discoveries of how it all works, and to the information that I posted at the beginning of this thread.
    I had my private server, and I was attempting to transfer account data from an old server, to a newer one I created. It pretty much came down to either all, or nothing. The inter-relational ties between the databases were just too complex for me to separate the data without corrupting it and making it so that either the character was unplayable, the account was unplayable, or crashing the server entirely.

    When the two servers share the same databases, it would probably be easier, but still will be too difficult of a task to be automated reliably. I was able to do it successfully in most cases when they ran off the same databases, but it only had about a 90% success rate.

    How would inter-relational ties matter when you transfer from server A to B? Just discard them all. You won't have the same friend list or guild or spouse.

    All that matters is gear, and I don't think gears have any inter-relational ties?

    And people can delete a character without the server crashing? So how would the data corrupt if you remove it from server A?
  • Zoldi - Morai
    Zoldi - Morai Posts: 1,595 Arc User
    edited June 2015
    unrefuted wrote: »
    How would inter-relational ties matter when you transfer from server A to B? Just discard them all. You won't have the same friend list or guild or spouse.

    All that matters is gear, and I don't think gears have any inter-relational ties?

    And people can delete a character without the server crashing? So how would the data corrupt if you remove it from server A?

    You're totally right and PWE knows how to do that since some people got their account "transferred" actually even if this is a taboo topic. The only thing that would matter is if the name is already taken, but as when you create a new toon. Usually people that got their account transferred, changed their name anyway. Then you give them the gears (with no name on it), with refines and gemms, the skills, the level (xp) and the cultivation.
    No need to deal with most relations between tables, 'cause you don't care about history.
    ~ Sage Mystic 105-105-105 ~
    ~ Sage Sin 103-103-102 ~
    ~ Sage vita Barb 103-102-101 ~
    ~ Sage BM 102-101 ~ Demon Archer 102 ~
    ~ Sage Cleric 102 ~ Demon Wizard 101 ~
    ~ Sage Seeker 101 ~ Sage Psy 101 ~
    ~ Sage Heavy Veno 101 ~ Demon Storm 9X ~ Dusk 6x ~
    Started playing this game in 2007 on Oracle (PW-MY) : Demon MG 101 - Sage EP 99 - WF 89
  • DEMHEALSMAN - Dreamweaver
    DEMHEALSMAN - Dreamweaver Posts: 1,396 Arc User
    edited June 2015
    unrefuted wrote: »
    How would inter-relational ties matter when you transfer from server A to B? Just discard them all. You won't have the same friend list or guild or spouse.

    All that matters is gear, and I don't think gears have any inter-relational ties?

    And people can delete a character without the server crashing? So how would the data corrupt if you remove it from server A?
    Because transferring isn't the same as merging.
    Soon™
    Well, maybe later, semi-retired.
  • Asterelle - Sanctuary
    Asterelle - Sanctuary Posts: 930 Arc User
    edited June 2015
    Because transferring isn't the same as merging.

    A merge is just transferring every account.
    [sigpic][/sigpic]
    PWI Calculators - aster.ohmydays.net/pw
  • DEMHEALSMAN - Dreamweaver
    DEMHEALSMAN - Dreamweaver Posts: 1,396 Arc User
    edited June 2015


    A merge is just transferring every account.

    Merge is not the same as transfer.
    Merge needs information from 2 places to be combined.
    Transfer would need to just place the data from point A to point B.

    Anyway I guess it's how you look at it, in PWI's case just transferring data from 1 server to the other should suffice, granted there wasn't the issue of names, which can be resolved by adding a prefix/postfix.
    People just wouldn't be happy about losing friendlist, marriage, faction, etc.

    But that is just my opinion and interpretation of both words.
    Soon™
    Well, maybe later, semi-retired.
  • unrefuted
    unrefuted Posts: 107 Arc User
    edited June 2015
    Because transferring isn't the same as merging.

    Was talking about transferring, not merging. Check the post I quoted.
  • opkossy
    opkossy Posts: 11,177 Community Moderator
    edited June 2015
    Merge is not the same as transfer.
    Merge needs information from 2 places to be combined.
    Transfer would need to just place the data from point A to point B.
    Yeah um... a transfer is, quite literally, doing most of the same things that you'd do for a merge but with fewer accounts.

    The closest things to "transfers" we've had weren't real transfers so much as them creating a new character and using tools available to have it match another one. They've never actually performed a real transfer of a character from server A to server B.
    (Insert fancy image here)
    image
  • RankNine - Momaganon
    RankNine - Momaganon Posts: 1,241 Arc User
    edited June 2015
    Merge is not the same as transfer.
    Merge needs information from 2 places to be combined.
    Transfer would need to just place the data from point A to point B.

    according to your definition, the EU-Merge was actually a giant transfer.
    [SIGPIC][/SIGPIC]
    noob, can only run, spawnkiller, only white vodoo, only plays for kd,
    never kills anyone, only gear, no skill, no life, cash only, eats dogfood to cash more, lives at moms,
    only survives cause of cleric heals, if we had your gear you would lose. b:cryb:cryb:cry
  • opkossy
    opkossy Posts: 11,177 Community Moderator
    edited June 2015
    Talking to me and Ast or to the guy you quoted? I ask because people quote folks they aren't responding to often enough as is. b:chuckle
    (Insert fancy image here)
    image
  • DEMHEALSMAN - Dreamweaver
    DEMHEALSMAN - Dreamweaver Posts: 1,396 Arc User
    edited June 2015
    according to your definition, the EU-Merge was actually a giant transfer.

    Which it was, in my opinion
    Merge = Point A ==> Point B <== Point C
    Transfer = Point A ==> Point B

    That is how I see both definitions, but this thread isn't about what my definition of the words are.
    One could also say that Point A was merged into Point B.

    Regardless, in the situation of a merge, in case the same data is found twice, you have to make a decision; what will be the changes for either or both point A or C data? Will you just rename A? will you just rename C? will you rename both?

    In the transfer situation you only have to think about what you are going to rename the data of Point A to in case there are duplicates.

    Again, that's just how I see the difference between the 2.

    Point is, neither one is easy, but a transfer imo is generally less of a hassle than a complete fullfledge merge.
    Soon™
    Well, maybe later, semi-retired.
  • ZentDreigon - Raging Tide
    ZentDreigon - Raging Tide Posts: 456 Arc User
    edited June 2015
    Transfer implies removing something from somewhere and putting it in another place, much like cut+paste. Merge implies combining two things so they don't get messed up during the transfer from A to B or A and B to C.

    Yes, a transfer is a merge. No one in their sanity would cut data from one server to paste it in another. Would be a copy+paste thing.

    Where I work we had to merge two databases which had conflicting data. It wasn't just a matter of taking from A and putting into B.



    Basically...

    Transfer: Majin Boo eats chocolate people and "transfers" them into his pink self.
    Merge: Goku and Vegeta use the fusion to combine themselves into one awesome thing.

    b:avoid From work to anime in one post.
    [SIGPIC][/SIGPIC]
    > A demon is driven by desire, pure desire, that cannot be stopped by reason or logic. - Mo Zun
    > Believing in demons doesn't mean believing in evil. Demons are not necessarily an evil thing. - Chin Wuming
  • XXxICExXx - Morai
    XXxICExXx - Morai Posts: 145 Arc User
    edited June 2015
    Rather than merge any into an existing one wouldn't it be easier, (that is if PWI/PWE had any inclination to achieve this) to create a whole new server and map all the merging details to this? Keep it a test server until all the data was properly transferred/merged and then allow those that existed in those servers to log in.

    It would essentially be a duplicate and those servers that were being merged could still log into the existing servers and play they characters until such time as the duplicate was properly mapped and tested?

    They wouldn't have had this issue if they had used a centralized cluster server like a well known Space MMO uses. b:shutup
  • Barbski - Archosaur
    Barbski - Archosaur Posts: 76 Arc User
    edited June 2015
    you might be right on some technicalities, but I completely disagree with your overall analysis. servers that can merge should. it's not that we don't have enough players, it's that we're spread too thin and can't positively impact each other's gaming experience. what we see today is the result of too many servers.

    it's where players are in relation to each other that matters. its the people on your server, fighting with or against you, that frame your desire - or lack thereof - to play; even if inadvertently. those on other servers, though? well, hell, they might as well not even exist. in a sense, they don't.

    more capable factions = more competition. more competition = intrigue + excitement = a server dynamic conducive for more overall desire to play from the community = impetus for more support / incentivized CS'ing. would you spend money to play on a server dominated by 1 faction, or 4 factions vying for power? it doesn't matter if g16's find it harder to get on the map.
  • Reliea - Sanctuary
    Reliea - Sanctuary Posts: 685 Arc User
    edited June 2015
    I can only speak from a Sanct perspective, because that's where I stick to.
    Population and activity levels by my criteria (amount of players visible in active areas, amount and variety of dungeons being advertised in wc, catshop population) haven't reached levels where I feel the server is in any danger or could be positively impacted by a merge. There are slow times, but that's a time-zone issue more than a flat population issue.

    A merge is not going to stop people from leaving the game.
    Main reasons why people quit PWI:
    1) End game player in OP gear bored 1 shotting people over n over
    2) End game player in low/average gear frustrated over their lack of OPness with no desire to work for or pay for better gear
    3) Player fed up with the mismanagement of game, be it exploit, bad update, unfixed glitches, horrible customer service etc
    4) Player interpersonal drama (aka relationships gone bad)
    None of those things will be effected by a merger in any positive way, and at least 1 and 2 could be exacerbated by an influx of players to an established server ecosystem.
    A merge is also not going to solve reason number 3, and seeing how obviously difficult it would be to execute, I could predict that the problems caused by one might cause a greater exodus than if they had just left well enough alone.

    NW is not a fair judge of population as clearly more people opt out these days.
    TW is a non-issue, as after factions re-establish themselves it will revert quickly back to the typical 2-3 factions with the largest willing participant population with the "best" gear taking over. The only interesting part of it will be seeing how native factions fair against the merged factions, and that will only last so long before they too merge and break apart as factions do.

    Considering they can't just suddenly decide to merge the old servers, you can assume that if they were going to attempt it, it would take a good 1-2 years of planning before execution. They can't even do a forum move without epic flaws, and we somehow expect them to be able to merge servers? XD How would you feel if they did do it, and somehow effed up your beloved characters you've spent so much time/money building up? Is it worth the risk? Me personally, I say no. I feel it's akin to putting my luggage in the hands of an airline with dodgy technical prowess. I might get there, but my bags are gone forever, only in-game, it is probably not worth rebuilding everything lost since it took me years to get to this point.

    Can they do it? Possibly.
    Should they do it? For some servers, maybe.
    Will they do it? Probably not.
    Why? The time it takes, the effort it takes, the money it takes, the risks entailed.
    If they are going to do it, they should start thinking about it now, because if they even do, it will still take a long time to get to the point that they can.
    But the answer really comes from an honest look at the attention they pay to this game now, and that's how I conclude there will never be a server merger, ever ever. If I'm wrong, my head will probably explode from sheer surprise.
  • Auldwulf - Dreamweaver
    Auldwulf - Dreamweaver Posts: 109 Arc User
    edited June 2015
    I see more people now, than ever on Dreamweaver. Places that used to be virtually empty in Arch, like the north section, now crowded, players more than ever, afk, leaving their toons all over important npcs, world chat with much more degenerative behavior than ever, on and on.

    Would I like to see a server merge, and possibly triple, or even quadruple these things? My answer is simply no....just no.

    The community has really gotten bad, nothing like it used to be. I just think it make for a much less desirable playing environment merging severs, than it already is. As far as more competition? People will still deal with the same issues, just more of it, so no thank you. If it's not broke, don't fix it.
  • DEMHEALSMAN - Dreamweaver
    DEMHEALSMAN - Dreamweaver Posts: 1,396 Arc User
    edited June 2015
    I see more people now, than ever on Dreamweaver. Places that used to be virtually empty in Arch, like the north section, now crowded, players more than ever, afk, leaving their toons all over important npcs, world chat with much more degenerative behavior than ever, on and on.

    Are we playing on the same server? Lately i've seen nothing but a degrading activity on dreamweaver. It's always the exact same people doing the exact same thing at the exact same time. For example FSP and BH forming.
    Soon™
    Well, maybe later, semi-retired.
  • Juliettees - Lost City
    Juliettees - Lost City Posts: 1 Arc User
    edited June 2015
    Since there were many entries I skipped some but let me clean the dust in the air a bit.

    Anyone knows a bit of PWI's past attendances can tell you that those old servers handled x10 of the presently online years ago. The technology has been updated/upgraded since then so what you were expecting to take a long time merging the server regardless of document complexity can be done at most in a day with a "group of professionals" who knows what they are dealing with and decent servers and well coded server-sided applications.

    Yet we lack that "professionals" as this company sucks badly in all the way from support to game patches, rewards, events etc so we can face at most 2 days of shut servers and in case of a fail with a rollback bringing the old system.

    I don't care about NW/TW etcetc attendances since I have seen west arch where there was not even one spot to step on with catshops overcrowd 3-4 years ago

    There are literally 2 options for us ahead:
    1. Letting the game die in upcoming 1-2 years
    2. Merging the servers to provide a better in-game experience to keep old players.

    Those databases you mention have the info on latest login date/time where you can figure your priorities to move tunes in to the merged server. The overly crowded numbers seen as getting online must be divided at least 4 or 5 because of alt farming. Most of those alts have nearly no data to be moved which also decreases the workload since they are only used for bh/fs etc.

    For the obvious reason the merge can not be done as you think "by hand" but by a well coded cross server application that should be previously tested on multiple test servers since any human participation will increase the fail possibility.

    There we are facing the real problem that the company providing this servers to us doesn't have the right to touch any of the code either server-side or our local. They don't have the coders/engineers that have any clue on what's going on under the hood. So to clear, PWI can not handle this but only PWCN can. Ok lets give it a little consideration on what we are getting as updates and how many errors/bugs/fails we have seen till today. PWCN side also sucks on coding.

    So in these terms and with these "professionals", merge has NO GO.

    There are many points more to show but needless to tell they will fail if they try a merge so better keep our games and tunes.
  • GingerSprite - Sanctuary
    GingerSprite - Sanctuary Posts: 15 Arc User
    edited June 2015
    I think the biggest issue with merging server is the load on the hardware. But then, if the two servers have not a lot of load because of low population, maybe not so bad?

    Three things we can be sure of thogh

    1. It's not gunna happen any time soon.
    2. Only a couple of servers have a realistic option of being merged.
    3. If it does happen, PWE will **** it up.