From b74affcd518fb3a4c18797c37e1bf0864b36c284 Mon Sep 17 00:00:00 2001
From: sbosse <sbosse@uni-bremen.de>
Date: Mon, 21 Jul 2025 22:48:04 +0200
Subject: [PATCH] Mon 21 Jul 22:43:21 CEST 2025

---
 js/dos/dns_srv.js | 2169 +++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 2169 insertions(+)
 create mode 100644 js/dos/dns_srv.js

diff --git a/js/dos/dns_srv.js b/js/dos/dns_srv.js
new file mode 100644
index 0000000..87eb757
--- /dev/null
+++ b/js/dos/dns_srv.js
@@ -0,0 +1,2169 @@
+/**
+ **      ==============================
+ **       O           O      O   OOOO
+ **       O           O     O O  O   O
+ **       O           O     O O  O   O
+ **       OOOO   OOOO O     OOO  OOOO
+ **       O   O       O    O   O O   O
+ **       O   O       O    O   O O   O
+ **       OOOO        OOOO O   O OOOO
+ **      ==============================
+ **      BSSLAB, Dr. Stefan Bosse http://www.bsslab.de
+ **
+ **      COPYRIGHT: THIS SOFTWARE, EXECUTABLE AND SOURCE CODE IS OWNED
+ **                 BY THE AUTHOR(S).
+ **                 THIS SOURCE CODE MAY NOT BE COPIED, EXTRACTED,
+ **                 MODIFIED, OR OTHERWISE USED IN A CONTEXT
+ **                 OUTSIDE OF THE SOFTWARE SYSTEM.
+ **
+ **    $AUTHORS:     Stefan Bosse
+ **    $INITIAL:     (C) 2015-2017 bLAB
+ **    $CREATED:     23-4-15 by sbosse.
+ **    $VERSION:     1.3.8
+ **
+ **    $INFO:
+ **      DNS Server with local file system storage and row capability cache.
+ *
+ *     $ENDINFO
+ */
+"use strict";
+
+var util = Require('util');
+var Io = Require('com/io');
+var Net = Require('dos/network');
+var Std = Require('dos/std');
+var Sch = Require('dos/scheduler');
+var Buf = Require('dos/buf');
+var Rpc = Require('dos/rpc');
+var Cs = Require('dos/capset');
+var Comp = Require('com/compat');
+var String = Comp.string;
+var Array = Comp.array;
+var Perv = Comp.pervasives;
+var Printf = Comp.printf;
+var div = Perv.div;
+var assert = Comp.assert;
+var Rand = Comp.random;
+var Status = Net.Status;
+var Command = Net.Command;
+var Rights = Net.Rights;
+var AfsInt = Require('dos/afs');
+var Afs = Require('dos/afs_srv_common');
+var Dns = Require('dos/dns_srv_common');
+var DnsInt = Require('dos/dns');
+var Fs = Require('fs');
+var Fcache = Require('dos/fcache');
+var Cache = Require('dos/cache');
+var Afs_file_state = Afs.Afs_file_state;
+
+var log = 0;
+var CACHETMO = 30;    // seconds
+
+/** DNS Dns_server
+ *
+ * @param {rpcint} rpc
+ * @constructor
+ * @typedef {{rpc,afs,std,cs,dns_super:dns_super,dns_part:string,dns_part_fd,cache_inode,cache_data,inode_cache_entry,
+ *            rootdir:dns_dir,fs_default,
+ *            block,to_block,of_block,
+ *            live_set,live_get,live_read,live_write,
+ *            read_inode,write_inode,sync_inode,
+ *            inode_of_obj,cap_of_obj,update_inode,
+ *            iocachebypass,sync_write_mode,
+ *            stats}} dns_server~obj
+ * @see dns_server~obj
+ * @see dns_server~meth
+ */
+var dns_server = function (rpc,options) {
+    this.verbose=options.verbose||0;
+    this.rpc=rpc;
+    this.afs=AfsInt.AfsInt(rpc);
+    this.std=Std.StdInt(rpc);
+    this.cs=Cs.CsInt(rpc);
+    /**
+     *
+     * @type {dns_super|undefined}
+     */
+    this.dns_super=undefined;
+    this.dns_part=''; // DNS inode partition file path
+    this.dns_part_fd=undefined;
+
+    /**
+     *
+     * @type {fsc_cache|undefined}
+     */
+    this.cache_inode = undefined;
+    /**
+     *
+     * @type {cache|undefined}
+     */
+    this.cache_data = undefined;
+    this.inode_cache_entry=undefined;
+
+
+    /** Capability cache to speed-up row
+     *  capability restriction using a column mask.
+     *  Key: cap-key+mask
+     *  Value: Restricted cap
+     *
+     */
+    this.cache_cap = Cache.Cache(1000);
+    
+    
+    this.rootdir=undefined;
+    this.fs_default=0;
+
+    this.block=function() {};
+    this.to_block=function() {};
+    this.of_block=function() {};
+
+    /**
+     *
+     * @param {number} obj
+     * @param {number} time
+     * @param {number} flag
+     */
+    this.live_set=function (obj,time,flag) {};
+    /**
+     *
+     * @param {number} obj
+     * @returns {{time: number, flag: number}}
+     */
+    this.live_get=function (obj) {return {time:0,flag:0}};
+    /**
+     *
+     * @returns {(Status.STD_OK|*)}
+     */
+    this.live_read=function () {return 0;};
+    /**
+     *
+     * @returns {(Status.STD_OK|*)}
+     */
+    this.live_write=function () {return 0;};
+
+    /**
+     *
+     * @param obj
+     * @param addr
+     * @param data
+     * @param size
+     * @returns {number}
+     */
+    this.read_inode=function(obj,addr,data,size) {return 0};
+    /**
+     *
+     * @param obj
+     * @param addr
+     * @param data
+     * @param size
+     * @returns {number}
+     */
+    this.write_inode=function(obj,addr,data,size) {return 0};
+    this.sync_inode=function() {};
+
+    /**
+     *
+     * @param {number} obj
+     * @returns {{stat:(Status.STD_OK|*),inode:(dns_inode|undefined)}}
+     */
+    this.inode_of_obj=function(obj){return {stat:0,inode:undefined}};
+    /**
+     *
+     * @param {number} obj
+     * @returns {{stat:(Status.STD_OK|*),cap:(capability|undefined),inode:(dns_inode|undefined)}}
+     */
+    this.cap_of_obj=function(obj){return {stat:0,cap:undefined,inode:undefined}};
+    this.update_inode=function(){};
+
+    this.iocache_bypass=false;
+    this.sync_write_mode=false;
+
+    this.stats = {
+        op_read:0,
+        op_modify:0,
+        op_create:0,
+        op_destroy:0,
+        op_touch:0,
+        op_age:0
+    }
+};
+
+/** Initialite and start up services
+ *
+ */
+
+dns_server.prototype.init = function () {
+  var self=this;
+  Sch.AddTimer(1000,'DNS Server Garbage Collector',function () {
+    self.cache_cap.refresh(function (key,data) {
+      if (data.tmo) { data.tmo--; return data.tmo>0}
+      else return true;      
+    });
+  });
+}
+
+/** Get the current system time in 10s units
+ *
+ * @returns {number}
+ */
+dns_server.prototype.time =function () {
+    var self=this;
+    return div(Perv.time(),10);
+};
+
+
+/** Acquire and lock a directory [BLOCKING]
+ *
+ * @param {number} obj
+ * @param {function((Status.STD_OK|*),(dns_dir|undefined))} callback
+ */
+dns_server.prototype.acquire_dir = function (obj,callback) {
+    var self=this;
+    var dir=undefined;
+    var stat;
+    Sch.ScheduleBlock([
+        function () {self.dns_super.lock();},
+        function () {
+            var res;
+            res=self.read_dir(obj,function (_stat,_dir) {
+                stat=_stat;
+                dir=_dir;
+            });
+            if (stat==Status.STD_OK) {
+                dir.lock();
+            }
+        },
+        function () {
+            self.dns_super.unlock();
+            Io.log((log < 1) || ('Dns_srv.acquire_dir: done ' + obj));
+            callback(stat,dir);
+        }
+    ],function (e) {
+        if (typeof e == 'number') callback(e,undefined); else {
+            Io.printstack(e,'Dns_server.acquire_dir');
+            callback(Status.STD_SYSERR,undefined);
+        }
+    });
+};
+
+/** Release an acquired directory. Flush all pending writes - the
+ ** super structure and the i-node table, and the directory itself
+ ** if modified. Return the status of the operation. [BLOCKING]
+ *
+ *
+ * @param {dns_dir} dir
+ * @param {function(Status.STD_OK|*)} callback
+ */
+dns_server.prototype.release_dir = function (dir,callback) {
+    var stat=Status.STD_OK;
+    Io.log((log < 1) || ('dns_srv.release_dir: ' + util.inspect(dir)));
+    if (dir != undefined) {
+        if ((dir.dd_state == Dns.Dns_dir_state.DD_unlocked ||
+             dir.dd_state == Dns.Dns_dir_state.DD_modified)) {
+            this.modify_dir(dir, function (_stat) {
+                callback(_stat);
+            });
+        }
+        else {
+            stat = Status.STD_OK;
+            dir.unlock();
+            if (callback) callback(stat);
+        }
+    } else {
+        stat=Status.STD_ARGBAD;
+        if (callback!=undefined) callback(stat);
+    }
+};
+/** Release an acquired but unmodified directory. Flush all pending writes - the
+ ** super structure and the i-node table.
+ ** Return the status of the operation.
+ *
+ *
+ * @param {dns_dir} dir
+ * @returns {(Status.STD_OK|*)}
+ *
+ * @param dir
+ */
+dns_server.prototype.release_unmodified_dir = function (dir) {
+    var stat=Status.STD_OK;
+    Io.log((log < 1) || ('dns_srv.release_unmodified_dir: ' + util.inspect(dir)));
+    if (dir != undefined) {
+        if ((dir.dd_state == Dns.Dns_dir_state.DD_unlocked ||
+            dir.dd_state == Dns.Dns_dir_state.DD_modified)) {
+            Io.out('Dns_srv.release_unmodified_dir: Directory '+dir.dd_objnum+' was modified!');
+            return Status.STD_SYSERR;
+        }
+        else {
+            stat = Status.STD_OK;
+            dir.unlock();
+            return stat;
+        }
+    } else {
+        return Status.STD_ARGBAD;
+    }
+};
+
+
+/** Lookup a directory object (priv.prv_obj) and check the required rights.
+ *  The directory is returned unlocked!
+ *
+ * @param {privat} priv
+ * @param {number} req
+ * @param {function((Status.STD_OK|*),(dns_dir|undefined))} callback
+ */
+dns_server.prototype.lookup_dir = function(priv,req,callback) {
+    var self = this;
+    var obj = Net.prv_number(priv);
+    var stat;
+    var dir;
+
+    if (obj < 1 || obj >= this.dirs_top) {
+        callback(Status.STD_ARGBAD,undefined);
+    } else Sch.ScheduleBlock([
+        function () {
+            Io.log((log<1)||('Dns_server.lookup_dir: read_dir('+obj+')'));
+            self.read_dir(obj,function (_stat,_dir) {
+                stat=_stat;
+                dir=_dir;
+            })
+        },
+        function () {
+            if (stat!=Status.STD_OK) throw stat;
+            if (!Net.prv_decode(priv, dir.dd_random)) {
+                throw Status.STD_DENIED;
+            }
+            /*
+             ** When checking the presence of column rights, only take the
+             ** *actual*
+             ** columns present into acount (i.e., do not use dns_COLMASK here)
+             */
+            var ncols=dir.dd_ncols;
+            var rights = Net.prv_rights(priv);
+            var colbits = Dns.dns_col_bits[ncols];
+            var colmask = colbits-1;
+
+            Io.log((log<1)||('Dns_server.lookup_dir: '+Net.Print.private(priv))+' req='+req+' colmask='+colmask+' rights='+rights);
+
+            if ((rights & colmask) == 0 ||
+                (rights & req) != req) {
+                throw Status.STD_DENIED;
+            }
+            callback(Status.STD_OK,dir);
+        }
+    ],function (e) {
+        if (typeof e == 'number') callback(e,undefined); else {
+            Io.printstack(e,'Dns_server.lookup_dir');
+            callback(Status.STD_SYSERR,undefined);
+        }
+    });
+
+};
+
+/** A client request arrived.  Enter the critical section. Check the capability.
+ ** If it is correct, check whether the required rights are present or not.
+ ** Return the directory structure, but only if the required rights are present.
+ *
+ * @param priv
+ * @param req
+ * @param callback
+ */
+dns_server.prototype.request_dir = function(priv,req,callback) {
+    var self = this;
+    var obj = Net.prv_number(priv);
+    var stat;
+    var dir;
+
+    if (obj < 1 || obj >= this.dirs_top) {
+        callback(Status.STD_ARGBAD,undefined);
+    } else Sch.ScheduleBlock([
+        function () {
+            Io.log((log<1)||('Dns_server.request_dir: acquire_dir('+obj+')'));
+            self.acquire_dir(obj,function (_stat,_dir) {
+                stat=_stat;
+                dir=_dir;
+            })
+        },
+        function () {
+            if (stat!=Status.STD_OK) throw stat;
+            if (!Net.prv_decode(priv, dir.dd_random)) {
+                self.release_unmodified_dir(dir);
+                throw Status.STD_DENIED;
+            }
+            /*
+             ** When checking the presence of column rights, only take the
+             ** *actual*
+             ** columns present into acount (i.e., do not use dns_COLMASK here)
+             */
+            var ncols=dir.dd_ncols;
+            var rights = Net.prv_rights(priv);
+            var colbits = Dns.dns_col_bits[ncols];
+            var colmask = colbits-1;
+
+            Io.log((log<1)||('Dns_server.request_dir: '+Net.Print.private(priv))+' req='+req+' colmask='+colmask+' rights='+rights);
+
+            if ((rights & colmask) == 0 ||
+                (rights & req) != req) {
+                self.release_unmodified_dir(dir);
+                throw Status.STD_DENIED;
+            }
+            callback(Status.STD_OK,dir);
+        }
+    ],function (e) {
+        if (typeof e == 'number') callback(e,undefined); else {
+            Io.printstack(e,'Dns_server.request_dir');
+            callback(Status.STD_SYSERR,undefined);
+        }
+    });
+};
+
+/**
+ ** With a given directory capability set check if this directory belongs
+ ** to this server.
+ *
+ *
+ * @param {capset} dir
+ * @returns {private|undefined}
+ */
+dns_server.prototype.check_dir = function(dir) {
+    var self=this;
+    Io.log((log<1)||('Dns_server.check_dir: '+Cs.Print.capset(dir)));
+    if (dir==undefined) return undefined;
+    for (var i = 0; i<dir.cs_final;i++) {
+        var s = dir.cs_suite[i];
+        if (s.s_current == true) {
+            var cap = s.s_object;
+            if (Net.Equal.port(cap.cap_port, self.dns_super.dns_putport) == true &&
+                Net.prv_number(cap.cap_priv) != 0) {
+                /*
+                 ** A capability for me. Check it.
+                 */
+                return cap.cap_priv;
+            }
+        }
+    }
+    Io.log((log<1)||'Dns_server.check_dir: not for me');
+    return undefined;
+};
+
+
+/** Calculate the size of a directory structure.
+ *
+ * @param {dns_dir} dir
+ * @returns {number}
+ */
+dns_server.prototype.dir_size = function (dir) {
+    return dir.size();
+};
+
+/** Return a free objnum in the directory table (index).
+ ** Note: protect this function with the server lock untill
+ ** the directory creation is finished.
+ *
+ * @returns {number}
+ */
+dns_server.prototype.get_freeobjnum = function () {
+    var self=this;
+    var sup=this.dns_super;
+    var obj=-1;
+    Array.match(sup.dns_freeobjnums,
+        function(hd,tl){
+            sup.dns_freeobjnums=tl;
+            obj=hd;
+        },
+        function () {
+            obj=sup.dns_nextfree;
+            if((obj+1)==sup.dns_ndirs) {
+                Io.out('[DNS] Out of directory slots.');
+                throw Status.STD_NOSPACE;
+            }
+            sup.dns_nextfree=obj+1;
+        });
+    return obj;
+};
+
+/** Convert a directory structure into a capset with maybe
+ ** restricted rights.
+ *
+ * @param {dns_dir} dir
+ * @returns {capset}
+ */
+dns_server.prototype.capset_of_dir = function(dir,rights) {
+    var self=this;
+    var sup=this.dns_super;
+    var cs;
+    var cap = Net.Capability(
+        sup.dns_putport,
+        Net.prv_encode(dir.dd_objnum,rights,dir.dd_random)
+    );
+    cs=this.cs.cs_singleton(cap);
+    return cs;
+};
+
+/** Restrict a directory capability set.
+ *  The directoty structure from capability set 'dir' must be unlocked to avoid.
+ *
+ * @param {capset} cs
+ * @param mask
+ * @param {function((Status.STD_OK|*),capset|undefined)} callback
+ */
+dns_server.prototype.restrict = function(cs,mask,callback) {
+    var self=this,
+        cs=Cs.Copy.capset(cs),
+        stat=Status.STD_CAPBAD,
+        index=0;
+    Sch.ScheduleLoop(function() {
+        return (stat!=Status.STD_OK && index<cs.cs_final);
+    },[
+        function () {
+            var s = cs.cs_suite[index],
+                cap,key,cache;
+            index++;
+            if (s.s_current == true) {
+                cap = s.s_object;
+                
+                if (Net.Equal.port(cap.cap_port, self.dns_super.dns_putport) == true &&
+                    Net.prv_number(cap.cap_priv) != 0) {
+                    /*
+                     ** A capability for me. Restrict it.
+                     */
+                    self.lookup_dir(cap.cap_priv,0,function (_stat,_dir) {
+                        if (_stat==Status.STD_OK) {
+                            cap.cap_priv=Net.prv_encode(cap.cap_priv.prv_obj,
+                                cap.cap_priv.prv_rights & mask,
+                                _dir.dd_random);
+                        }
+                        stat=_stat;
+                    });  // dir is unlocked, no release!
+                 } else {
+                    Io.log((log<1)||('Dns_server.restrict: doing std_restrict for '+Net.Print.capability(cap)+' with mask='+mask));
+                    // Try to use capability cache - but can be outdated
+                    key=Net.key(cap);
+                    cache=self.cache_cap.lookup(key);
+
+                    if (cache && cache.restrict && cache.restrict[mask]) {
+                      Net.Copy.private(cache.restrict[mask].cap_priv,cap.cap_priv);
+                      stat=cache.stat;
+                    } else self.std.std_restrict(cap,mask,function (_stat,_cap) {
+                        Io.log((log<1)||('Dns_server.restrict: std_restrict returned: '+Status.print(_stat)));
+                        stat=_stat;
+                        if (stat==Status.STD_OK) {
+                          if (cache && cache.restrict) {
+                            cache.stat=stat; 
+                            cache.restrict[mask]=_cap;
+                            cache.tmo=CACHETMO;
+                          } else
+                            self.cache_cap.add(key,{restrict:self.cache_cap.map(mask,_cap),
+                                                    tmo:CACHETMO,stat:stat});
+                          Net.Copy.private(_cap.cap_priv,cap.cap_priv);
+                        } else {
+                          if (cache && cache.restrict) {
+                            cache.stat=stat; 
+                            cache.restrict[mask]=Net.nilcap; 
+                            cache.cap=cap
+                          } else 
+                            self.cache_cap.add(key,{restrict:self.cache_cap.map(mask,Net.nilcap),
+                                                    tmo:CACHETMO,stat:stat,cap:cap});
+                        }
+                    });
+                }
+            }
+        }
+    ],[
+        function () {
+            callback(stat,cs);
+        }
+    ],function (e) {
+        if (typeof e != 'number') {Io.printstack(e,'Dns_server.restrict');}
+        if (typeof e == 'number') callback(e,undefined); else callback(Status.STD_SYSERR,undefined);
+    });
+};
+/** Get defualt file server.
+ *
+ *
+ *
+ * @param {function((Status.STD_OK|*),(Dns_mode.DNSMODE_ONECOPY|*))} callback
+ */
+dns_server.prototype.get_def_fs = function (callback) {
+    var self=this;
+    var sup = this.dns_super;
+    var fs = sup.dns_fs;
+    var cap = undefined;
+    var stat;
+    var goods = 0;
+    var goodcap = undefined;
+    var mode;
+
+    var index = 0;
+    Sch.ScheduleLoop(
+        function() {
+            return index<2;
+        }, [
+            function () {
+                cap = fs.fs_cap[index];
+                if (cap != undefined) {
+                    self.std.std_info(cap,function (_stat,_info) {
+                        stat=_stat;
+                    })
+                } else stat=Status.STD_CAPBAD;
+            },
+            function () {
+                if (cap!=undefined) {
+                    if (stat == Status.STD_OK) {
+                        fs.fs_state[index] = Dns.Fs_state.FS_up;
+                        goods++;
+                        if (goodcap==undefined) {
+                            goodcap=cap;
+                            fs.fs_default=index;
+                        }
+                    }
+                    else
+                        fs.fs_state[index] = Dns.Fs_state.FS_down;
+                }
+            }
+        ],[
+            function () {
+                if (goodcap==undefined) stat=Status.STD_NOTFOUND;
+                if (goods==1) mode = Dns.Dns_mode.DNSMODE_ONECOPY;
+                if (goods==2) mode = Dns.Dns_mode.DNSMODE_TWOCOPY;
+                callback(stat,mode);
+            }
+        ],
+        function (e) {
+            if (typeof e == 'number') callback(e,undefined); else {
+                Io.printstack(e,'Dns_server.get_def_fs');
+                callback(Status.STD_SYSERR,undefined);
+            }
+
+    })
+};
+
+
+/**
+ ** The create, read, modify and utility functions needed
+ ** for the Dns_server module.
+ **
+ ** Note:
+ ** After a directory is locked, any modification of this
+ ** directory leads to a new AFS file object and a new
+ ** directory capability!
+ ** In one copy mode, the old cap will be destroyed, after
+ ** the new one is committed and the i-node mapping was
+ ** updated.
+ */
+
+
+/** Read the directory content from an AFS file. [BLOCKING]
+*
+ */
+dns_server.prototype.read_dir_file = function (inode,callback) {
+    var self=this;
+    var stat,dir;
+    var fs_default = self.dns_super.dns_fs.fs_default;
+    var buf = Buf.Buffer();
+
+    Sch.ScheduleBlock([
+        function () {
+            assert((inode.i_size>0)||('Dns_srv.read_dir_file: inode.i_size is zero'));
+            Io.log((log<1)||('Dns_server.read_dir_file cap='+Net.Print.capability(inode.i_caps[fs_default])+' '+util.inspect(inode)));
+            self.afs.afs_read(inode.i_caps[fs_default],buf,0,inode.i_size, function (_stat) {
+                stat=_stat;
+                if (stat==Status.STD_OK) dir=Dns.buf_get_dir(buf); else dir=undefined;
+            })
+        },
+        function () {
+            callback(stat,dir);
+        }
+    ],function (e) {
+        if (typeof e == 'number') callback(e,undefined); else {
+            Io.printstack(e,'Dns_server.read_dir_file');
+            callback(Status.STD_SYSERR,undefined);
+        }
+    });
+
+};
+
+/** Read a directory specified with his object number (index) number.
+ * Either the directory is currently cached, or the directory is read from an AFS file.
+ *
+ * @param {number} obj
+ * @param {function((Status.STD_OK|*), dns_dir)} callback
+ */
+dns_server.prototype.read_dir = function (obj,callback) {
+    var self=this;
+    var stat,dir,inode;
+    this.stats.op_read++;
+    var si=this.inode_of_obj(obj);
+    stat=si.stat;inode=si.inode;
+    Io.log((log<1)||('Dns_server.read_dir obj='+obj));
+    if (stat!=Status.STD_OK) callback(stat,undefined);
+    else Sch.ScheduleBlock([
+        function () {
+            dir=self.cache_data.lookup(inode.i_objnum);
+            if(dir==undefined) stat=Status.STD_NOTFOUND; else stat=Status.STD_OK;
+        },
+        function () {
+            if (stat==Status.STD_NOTFOUND) self.read_dir_file(inode,function (_stat,_dir) {
+              stat = _stat;
+              dir = _dir;
+            })
+        },
+        function () {
+            if (stat==Status.STD_OK) {
+                var tf=self.live_get(dir.dd_objnum);
+                dir.dd_live=tf.time;
+            }
+            callback(stat,dir);
+        }
+    ],function (e) {
+        if (typeof e == 'number') callback(e,undefined); else {
+            Io.printstack(e,'Dns_server.read_dir');
+            callback(Status.STD_SYSERR,undefined);
+        }
+    });
+
+};
+
+/**
+ * Write an existing directory (only if modified).
+ *
+ * @param {dns_dir} dir
+ * @param {function(Status.STD_OK|*)} callback
+ */
+dns_server.prototype.modify_dir = function (dir,callback) {
+    var i;
+    var stat;
+    var self = this;
+    var oldcaps;
+    var cap = undefined;
+    var obj = dir.dd_objnum;
+    var inode,inode2,dirsize;
+    var fs = this.dns_super.dns_fs;
+    var fs_default=fs.fs_default;
+    var buf = Buf.Buffer();
+    var si=this.inode_of_obj(obj);
+    stat=si.stat;inode=si.inode;
+    Io.log((log<1)||('Dns_server.modify_dir dir='+util.inspect(dir)));
+    this.stats.op_modify++;
+    /*
+     ** Distinguish these two cases:
+     **  dd_state = DD_modified -> a new AFS object must be created
+     **  dd_state = DD_unlocked -> commit the AFS object
+     */
+    if (stat!=Status.STD_OK) callback(stat);
+    else
+    {
+        if (dir.dd_state==Dns.Dns_dir_state.DD_modified)
+            Sch.ScheduleBlock([
+                /*
+                 ** Create a new AFS object. If in two copy mode,
+                 ** the copy will be duplicated later.
+                 ** Destroy the old capabilities after this oepration
+                 ** succeeded.
+                 */
+                function () {
+                    oldcaps = Array.map(inode.i_caps,function (cap) {return Net.Copy.capability(cap)});
+                    dirsize=dir.size();
+                    Dns.buf_put_dir(buf,dir);
+                    self.afs.afs_create(fs.fs_cap[fs_default],buf,dirsize,Afs.Afs_commit_flag.AFS_SAFETY,function (_stat,_cap) {
+                        stat=_stat;
+                        cap=_cap;
+                    })
+                },
+                function () {
+                    if (stat!=Status.STD_OK) throw stat;
+                    dir.dd_state=Dns.Dns_dir_state.DD_locked;
+                    var caps=[Net.nilcap,Net.nilcap];
+                    caps[fs_default]=cap;
+                    inode2=Dns.Dns_inode(obj,caps,dirsize,Afs.Afs_file_state.FF_locked);
+                    Buf.buf_init(buf);
+                    Dns.buf_put_inode(buf,inode2);
+                    /*
+                     ** Write the inode data through the cache to disk ...
+                     */
+                    stat=self.cache_inode.cache_write(self.inode_cache_entry,buf,Dns.off_of_inode(obj),Dns.DNS_INODE_SIZE);
+                    if (stat!=Status.STD_OK) throw stat;
+                    i=0;
+                    Sch.ScheduleLoop(function(index) {
+                        i=index;
+                        return index<2;
+                    },[
+                        /*
+                         ** Destroy the old caps.
+                         */
+                        function () {
+                            if(!Net.Equal.capability(oldcaps[i],Net.nilcap)) {
+                                self.std.std_destroy(oldcaps[i],function (_stat) {
+                                    stat=_stat;
+                                })
+                            }
+                        }
+                    ],[
+                        function () {
+                            stat=Status.STD_OK;
+                        }
+                    ])
+                },
+                function () {
+                    callback(stat)
+                }
+            ],function (e) {
+                if (typeof e == 'number') callback(e); else {
+                    Io.printstack(e,'Dns_server.modify_dir');
+                    callback(Status.STD_SYSERR);
+                }
+            });
+        else // dir.dd_state = DD_unlocked
+            Sch.ScheduleBlock([
+                function () {
+                    dirsize=self.dir_size(dir);
+                    Dns.buf_put_dir(buf,dir);
+                    self.afs.afs_modify(inode.i_caps[fs_default],buf,dirsize,0,Afs.Afs_commit_flag.AFS_SAFETY,function (_stat,_cap) {
+                        stat=_stat;
+                        cap=_cap;
+                    })
+                },
+                function () {
+                    if (stat!=Status.STD_OK) throw stat;
+                    dir.dd_state=Dns.Dns_dir_state.DD_locked;
+                    inode.i_size=dirsize;
+                    inode.i_state=Afs.Afs_file_state.FF_locked;
+                    Buf.buf_init(buf);
+                    Dns.buf_put_inode(buf,inode);
+                    stat=self.cache_inode.cache_write(self.inode_cache_entry,buf,Dns.off_of_inode(obj),Dns.DNS_INODE_SIZE);
+                },
+                function () {
+                    callback(stat)
+                }
+            ],function (e) {
+                if (typeof e == 'number') callback(e); else {
+                    Io.printstack(e,'Dns_server.modify_dir');
+                    callback(Status.STD_SYSERR);
+                }
+            });
+    }
+};
+
+/** Insert a new created directory into the DNS table.
+ *
+ * @param {dns_dir} dir
+ * @param {function((Status.STD_OK|*))} callback
+ */
+dns_server.prototype.create_dir = function (dir,callback) {
+    var stat;
+    var self = this;
+    var cap = undefined;
+    var obj = dir.dd_objnum;
+    var fs = this.dns_super.dns_fs;
+    var fs_default=fs.fs_default;
+    var buf = Buf.Buffer();
+
+    this.stats.op_create++;
+    Sch.ScheduleBlock([
+        function () {
+            /*
+             ** Create a new AFS object. If in two copy mode,
+             ** the copy will be duplicated later.
+             */
+            self.afs.afs_create(fs.fs_cap[fs_default],buf,0,Afs.Afs_commit_flag.AFS_UNCOMMIT,function (_stat,_cap) {
+                stat=_stat;
+                cap=_cap;
+            })
+        },
+        function () {
+            if (stat!=Status.STD_OK) throw stat;
+            dir.dd_state=Dns.Dns_dir_state.DD_unlocked;
+            self.live_set(obj,Dns.DNS_MAX_LIVE,1);
+            /*
+             ** First read the inode from disk to update
+             ** the inode cache (reads/writes always full blocks!).
+             */
+            stat=self.cache_inode.cache_read(self.inode_cache_entry,buf,Dns.off_of_inode(obj),Dns.DEF_INODE_SIZE);
+
+            if (stat!=Status.STD_OK) throw stat;
+
+            var fs_caps = [Net.nilcap,Net.nilcap];
+            fs_caps[fs_default]=cap;
+            var inode = Dns.Dns_inode(obj,fs_caps,0,Afs.Afs_file_state.FF_unlocked);
+            Buf.buf_init(buf);
+            Dns.buf_put_inode(buf,inode);
+            stat=self.cache_inode.cache_write(self.inode_cache_entry,buf,Dns.off_of_inode(obj),Dns.DEF_INODE_SIZE);
+            self.cache_data.add(obj,dir);
+
+            callback(stat);
+        }
+    ],function (e) {
+        if (typeof e == 'number') callback(e); else {
+            Io.printstack(e,'Dns_server.create_dir');
+            callback(Status.STD_SYSERR);
+        }
+    })
+};
+
+/** Create a new directory. Return the directory structure and
+ ** the status returned by the server create function. The super structure
+ ** is already modified by this function. The new directory remains
+ ** locked.
+ *
+ * @param colnames
+ * @param {function ((Status.STD_OK|*),(dns_dir|undefined))} callback
+ */
+dns_server.prototype.alloc_dir = function (colnames,callback) {
+    var self=this;
+    var sup=this.dns_super;
+    Sch.ScheduleBlock([
+        function () {
+            sup.lock();
+        },
+        function () {
+            var obj=self.get_freeobjnum();
+            var dir = Dns.Dns_dir(obj,colnames.length,0,colnames,Net.uniqport(),[],
+                Dns.Dns_dir_state.DD_unlocked,self.time(),Dns.DNS_MAX_LIVE);
+            assert(dir.try_lock());
+            sup.unlock();
+            self.create_dir(dir,function (_stat) {
+                if (_stat==Status.STD_OK) callback(_stat,dir);
+                else callback(_stat,undefined);
+            })
+        }
+    ]);
+};
+
+
+/** Delete a directory. Destroy associated AFS objects.
+ *
+ * @param {dns_dir} dir
+ * @param {function(Status.STD_OK|*)} callback
+ */
+dns_server.prototype.delete_dir = function (dir,callback) {
+    var stat;
+    var i;
+    var self = this;
+    var cap = undefined;
+    var obj = dir.dd_objnum;
+    var fs = this.dns_super.dns_fs;
+    var fs_default=fs.fs_default;
+    var buf = Buf.Buffer();
+    var oldcaps;
+    /*
+     ** Get the i-node mapping
+     */
+    var inode;
+    this.stats.op_destroy++;
+    var si=this.inode_of_obj(obj);
+    stat=si.stat;inode=si.inode;
+    if (stat!=Status.STD_OK) callback(stat);
+    else {
+        self.live_set(obj,0,0);
+        oldcaps = Array.map(inode.i_caps,function (cap) {return Net.Copy.capability(cap)});
+        /*
+         ** Destroy first the old caps
+         */
+        i=0;
+        Sch.ScheduleLoop(function(index) {
+            i=index;
+            return index<2;
+        },[
+            /*
+             ** Destroy the old caps.
+             */
+            function () {
+                if(!Net.Equal.capability(oldcaps[i],Net.nilcap)) {
+                    self.std.std_destroy(oldcaps[i],function (_stat) {
+                        stat=_stat;
+                    })
+                }
+            }
+        ],[
+            function () {
+                stat=Status.STD_OK;
+                /*
+                ** Now destroy ourself!
+                ** Write the i-node data through the cache to disk ...
+                */
+                inode.i_state=Afs_file_state.FF_invalid;
+                inode.i_size=0;
+                inode.i_caps=[Net.nilcap,Net.nilcap];
+                Dns.buf_put_inode(buf,inode);
+                stat=self.cache_inode.cache_write(self.inode_cache_entry,buf,Dns.off_of_inode(obj),Dns.DNS_INODE_SIZE);
+                /*
+                 ** Invalidate cache entry.
+                 */
+                self.cache_data.invalidate(inode.i_objnum);
+                callback(stat);
+            }
+        ], function(e) {
+            if (typeof e == 'number') callback(e); else {
+                Io.printstack(e,'Dns_server.delete_dir');
+                callback(Status.STD_SYSERR);
+            }
+
+        })
+    }
+};
+dns_server.prototype.dir_modified = function (dir) {
+    if (dir.dd_state == Dns.Dns_dir_state.DD_locked)
+        dir.dd_state = Dns.Dns_dir_state.DD_modified;
+};
+
+/**
+ * Delete an i-node without destroying AFS objects.
+ *
+ * @param {number} obj
+ * @param {function(Status.STD_OK|*)} callback
+ */
+dns_server.prototype.delete_inode = function (obj,callback) {
+    var stat;
+    var self = this;
+    var buf = Buf.Buffer();
+    /*
+     ** Get the inode mapping
+     */
+    var inode;
+    this.stats.op_destroy++;
+    var si=this.inode_of_obj(obj);
+    stat=si.stat;inode=si.inode;
+    if (stat!=Status.STD_OK) callback(stat);
+    else {
+        /*
+         ** Destroy ourself without destroying AFS objects!
+         ** Write the i-node data through the cache to disk ...
+         */
+        inode.i_state=Afs_file_state.FF_invalid;
+        inode.i_size=0;
+        inode.i_caps=[Net.nilcap,Net.nilcap];
+        Dns.buf_put_inode(buf,inode);
+        stat=self.cache_inode.cache_write(self.inode_cache_entry,buf,Dns.off_of_inode(obj),Dns.DNS_INODE_SIZE);
+        /*
+         ** Invalidate cache entry.
+         */
+        self.cache_data.invalidate(inode.i_objnum);
+        callback(stat);
+    }
+};
+
+/**
+ * Create generic server statistics informations.
+ *
+ * @returns {string}
+ */
+dns_server.prototype.stat = function () {
+    var self=this;
+    var stats,res;
+    stats=self.stats;
+    res = 'DNS server status\n';
+    res = res + 'Inode cache statistics:\n' +
+            (self.cache_inode.cache_stat()) +
+    '\nServer statistics\n\n'+
+        'Read:   '+Printf.sprintf2([['%8d',stats.op_read]])+
+        ' Modify: '+Printf.sprintf2([['%8d',stats.op_modify]])+
+        ' Create: '+Printf.sprintf2([['%8d',stats.op_create]])+'\n'+
+        'Touch:  '+Printf.sprintf2([['%8d',stats.op_touch]])+
+        ' Age:    '+Printf.sprintf2([['%8d',stats.op_age]])+
+        ' Destroy:'+Printf.sprintf2([['%8d',stats.op_destroy]])+'\n';
+     return res;
+};
+
+/**
+ * Touch directory and AFS content table objects
+ *
+ * @param {dns_dir} dir
+ * @param {function(Status.STD_OK|*)} callback
+ */
+dns_server.prototype.touch = function (dir,callback) {
+    var self=this,
+        stat,
+        cap,inode,
+        i;
+    this.stats.op_touch++;
+    /*
+     ** Keep object internally alive.
+     */
+    dir.dd_live=Dns.DNS_MAX_LIVE;
+    self.live_set(dir.dd_objnum,Dns.DNS_MAX_LIVE,1);
+    /*
+     ** Keep object(s) externally alive.
+     */
+    var sci = self.cap_of_obj(dir.dd_objnum);
+    stat=sci.stat;inode=sci.inode;
+    i=0;
+    if (stat!=Status.STD_OK) callback(stat);
+    else Sch.ScheduleLoop(function(index) {
+        i=index;
+        return index < 2;
+    },[
+        function () {
+            cap=inode.i_caps[i];
+            if (!Net.Equal.capability(cap,Net.nilcap)) {
+                self.std.std_touch(cap,function (_stat) {
+                    stat=_stat;
+                })
+            }
+        }
+    ],[
+        function () {
+            callback(Status.STD_OK);
+        }
+    ],function(e) {
+        if (typeof e == 'number') callback(e); else {
+            Io.printstack(e,'Dns_server.touch');
+            callback(Status.STD_SYSERR);
+        }
+    });
+};
+
+/** Age all objects (used and unused inodes). Destroy directories
+ ** with live time equal zero (only valid files). Remember that
+ ** the root dirctory (obj=1) can't be aged or deleted.
+ *
+ * @param  {function(Status.STD_OK|*)} callback
+ */
+dns_server.prototype.age = function (callback) {
+    var self=this,
+        sup = this.dns_super,
+        stat=Status.STD_OK,
+        i=2,
+        lt,
+        dir,
+        gone=0,
+        aged=0,
+        scavenge=[];
+    this.stats.op_age++;
+    if (self.verbose) Io.out('[DNS] Aging ..');
+    for(i=2;i<sup.dns_ndirs;i++) {
+      lt = self.live_get(i);
+      if (lt.flag == 1 && lt.time > 1 ) {                    
+          self.live_set(i,lt.time-1,lt.flag);
+          aged++;
+      } else {
+          self.live_set(i,0,lt.flag);
+      }
+      if (lt.flag==1 && lt.time==0) scavenge.push(i);
+    }
+    if (scavenge.length) Sch.ScheduleLoop(function() {
+        return i < scavenge.length;
+    },[
+        function () {
+          dir=undefined;
+           self.acquire_dir(scavenge[i],function (_stat,_dir) {
+                  stat=_stat;
+                  dir=_dir;
+           });
+        },
+        function () {
+          if (stat==Status.STD_OK) {
+            self.delete_dir(dir,function (_stat) {
+                stat=_stat;
+                gone++;
+            })
+          } else if (stat!=Status.STD_NOTNOW) {
+            /*
+             ** Maybe already deleted AFS object.
+             */
+            self.delete_inode(scavenge[i],function (_stat) {
+                stat=_stat;
+                gone++;
+            })
+          }
+        },
+        function () {
+          i++;
+          if (dir!=undefined) self.release_dir(dir,function (_stat) {
+              stat=_stat;
+          });
+        }
+    ],[
+        function () {
+            // sup.unlock();
+            if (self.verbose) Io.out('[DNS] '+aged+' aged, '+gone+' directories(s) deleted.');
+            else if (gone) Io.out('[DNS] '+gone+' directories(s) deleted.');
+            callback(Status.STD_OK);
+        }
+    ],function(e) {
+        // sup.unlock();
+        if (typeof e == 'number') callback(e); else {
+            Io.printstack(e,'Dns_server.age');
+            callback(Status.STD_SYSERR);
+        }
+    });
+  else callback(stat);
+};
+
+/** Append a row to a directory.
+ *
+ * @param {dns_dir} dir
+ * @param {dns_row} row
+ */
+dns_server.prototype.append_row = function (dir,row) {
+    var self=this;
+    dir.dd_rows.push(row);
+    dir.dd_nrows++;
+    this.dir_modified(dir);
+};
+
+/** Append a row to a directory.
+ *
+ * @param {dns_dir} dir
+ * @param {string} oldname
+ * @param {string} newname
+ * @returns {(Status.STD_OK|*)}
+ */
+dns_server.prototype.rename_row = function (dir,oldname,newname) {
+    var self=this;
+    var row = Array.find(dir.dd_rows,function(_row) {
+       return String.equal(_row.dr_name,oldname);
+    });
+    if (row!=undefined) {
+        row.dr_name=newname;
+        self.dir_modified(dir);
+        return Status.STD_OK;
+    } else return Status.STD_NOTFOUND;
+};
+
+/** Delete a row of a directory.
+ *
+ * @param {dns_dir} dir
+ * @param {string} name
+ */
+dns_server.prototype.delete_row = function (dir,name) {
+    var self=this;
+    dir.dd_rows=Array.filter(dir.dd_rows,function (_row) {
+        return !(String.equal(_row.dr_name,name));
+    });
+    dir.dd_nrows=Array.length(dir.dd_rows);
+    this.dir_modified(dir);
+};
+
+/** Replace a row in a directory.
+ *
+ * @param {dns_dir} dir
+ * @param {string} name
+ * @param {capset} newcs
+ * @returns {(Status.STD_OK|*)}
+ */
+dns_server.prototype.replace_row = function (dir,name,newcs) {
+    var self=this;
+    var row = Array.find(dir.dd_rows,function(_row) {
+        return String.equal(_row.dr_name,oldname);
+    });
+    if (row!=undefined) {
+        row.dr_capset=newcs;
+        this.dir_modified(dir);
+    } else return Status.STD_NOTFOUND;
+};
+
+
+/** Search a row in a directory.
+ *
+ * @param {dns_dir} dir
+ * @param {string} name
+ * @returns {(dns_row|undefined)}
+ */
+dns_server.prototype.search_row = function (dir,name) {
+    var self=this;
+    var row = Array.find(dir.dd_rows,function(_row) {
+        return String.equal(_row.dr_name,name);
+    });
+    return row;
+};
+/**
+ * @returns {dns_super}
+ */
+dns_server.prototype.read_super = function () {
+    return this.dns_super;
+};
+
+
+/** Returns (used,lifetime) tuple. This functions first probes for an
+ ** used i-node, and in the case of a used i-node was found, it ages the live time.
+ ** If the lifetime reaches zero, the object must be destroyed.
+ ** (by this module).
+ *
+ *
+ * @param {number} obj
+ * @param {function(number,number)} callback  function(used,live)
+ * @returns {{used: boolean, live: number}}
+ */
+dns_server.prototype.age_obj = function (obj,callback) {
+    var used,live,dir;
+    this.stats.op_age++;
+    Sch.ScheduleBlock([
+        function () {
+            self.dns_super.lock();
+        },
+        function () {
+            var tf = self.live_get(obj);
+            if (tf.flag == 1 && tf.time>1) {
+                self.live_set(obj,tf.time-1,1);
+                used=true;live=tf.time-1;
+            } else if (tf.flag==1) {
+                self.live_set(obj,0,1);
+                used=true;live=0;
+            } else {
+                used=false;live=0;
+            }
+            self.dns_super.unlock();
+            callback (used,live);
+        }
+    ]);
+};
+
+/**
+ * Flush the caches (if any).
+ *
+ * @returns {(Status.STD_OK|*)}
+ */
+dns_server.prototype.sync = function () {
+    var stat;
+    // TODO
+    return stat;
+};
+
+/**
+ * Exit the server.
+ *
+ * @returns {(Status.STD_OK|*)}
+ */
+dns_server.prototype.exit = function () {
+    Io.out('[DNS] Exit. Writing live table...');
+    return this.live_write();
+};
+
+
+
+
+/** Create the directory filesystem (one partition = UNIX file). [BLOCKING]
+ **
+ ** Args:
+ **  label:      the filesystem label string [max 256 chars]
+ **  ninodes:    number of i-nodes ( = maximal number of directories)
+ **  cols:       Column name array (dns_colnames)
+ **  colmasks:   Column mask array (dns_generic_colmask)
+ **  part_inode:   Partition path and name
+ **  fs_caps:  AFS server(s). All directories are saved there. Up to
+ **              two file servers can be specified. The first is the default.
+ **
+ ** Return:
+ **  status
+ **  dns_super
+ *
+ *
+ * @param {string} label
+ * @param {number} ninodes
+ * @param {number} blocksize
+ * @param {string []} cols
+ * @param {number []} colmasks
+ * @param {string} part_inode
+ * @param {capability []} fs_caps
+ * @param {boolean} overwrite
+ * @param {function(Status.STD_OK|*)} callback
+ */
+dns_server.prototype.create_fs = function (label,ninodes,blocksize,cols,colmasks,part_inode,fs_caps,overwrite,callback) {
+    var self = this;
+    function create (callback) {
+        var i, n;
+
+        var stat = Status.STD_OK;
+        if (fs_caps==undefined || Array.empty(fs_caps) || Net.Equal.capability(fs_caps[0],Net.nilcap)) {
+            Io.out('[DNS] Cannot create DNS file system. No file server capability provided. Fatal.');
+            throw Status.STD_ARGBAD;
+        }
+        Io.out('[DNS] Creating DNS tree...');
+        Io.out('[DNS] Blocksize: ' + blocksize + ' [bytes]');
+        Io.out('[DNS] Number of total inodes (dirs): ' + ninodes);
+
+        /*
+         ** First create a server port and derive the public port.
+         */
+        var privport = Net.uniqport();
+        var pubport = Net.prv2pub(privport);
+        var checkfield = Net.uniqport();
+        var ncols = cols.length;
+
+
+        Io.out('[DNS] Private Port: ' + Net.Print.port(privport));
+        Io.out('[DNS] Public Port: ' + Net.Print.port(pubport));
+        Io.out('[DNS] Checkfield: ' + Net.Print.port(checkfield));
+
+        Io.out('[DNS] AFS server [1]: '+Net.Print.capability(fs_caps[0]) + ' (default)');
+        Io.out('[DNS] AFS server [2]: '+Net.Print.capability(fs_caps[1]));
+        self.fs_default=0;
+
+        if (!Io.exists(part_inode)) {
+            Io.out('[DNS] Creating directory partition ' + part_inode);
+        } else if (overwrite) {
+            Io.out('[DNS] Overwriting existing directory partition ' + part_inode);
+        } else {
+            Io.err('[DNS] Found existing directory partition ' + part_inode);
+        }
+        var part_fd = Io.open(part_inode, (overwrite ? 'w+' : 'r+'));
+        self.dns_part = part_inode;
+        self.dns_part_fd = part_fd;
+
+        /*
+         ** The partition magic headers. After the magic string,
+         ** the value 0xaa is written.
+         */
+        var magic1 = Buf.Buffer();
+        Buf.buf_put_string(magic1, Dns.MAGIC_STR);
+        Buf.buf_pad(magic1, Dns.DEF_MAGIC_SIZE, 0xaa);
+
+        Io.out('[DNS] Writing partition magic header... ');
+        n = Buf.buf_write(part_fd, magic1);
+        if (n != Dns.DEF_MAGIC_SIZE) throw Status.STD_IOERR;
+
+        /*
+         ** The disk super structure.
+         */
+
+        var super1 = Buf.Buffer();
+        Buf.buf_put_string(super1, label);
+        Buf.buf_pad(super1, 256, 0xaa);
+        Buf.buf_put_int32(super1, ninodes);
+        Buf.buf_put_int32(super1, blocksize);
+        Buf.buf_put_port(super1, privport);
+        Buf.buf_put_port(super1, pubport);
+        Buf.buf_put_port(super1, checkfield);
+        Buf.buf_put_int32(super1, ncols);
+
+        for (i = 0; i < ncols; i++) {
+            Buf.buf_put_string(super1, cols[i]);
+        }
+        for (i = 0; i < ncols; i++) {
+            DnsInt.buf_put_rights(super1, colmasks[i]);
+        }
+        for (i=0;i<2;i++) {
+            Buf.buf_put_cap(super1,fs_caps[i]);
+        }
+
+        Io.out('[DNS] Done. ');
+        Io.out('[DNS] Writing super block ... ');
+        Buf.buf_pad(super1, Dns.DEF_BLOCK_SIZE - Dns.DEF_MAGIC_SIZE, 0xaa);
+        n = Buf.buf_write(part_fd, super1);
+        if (n != (Dns.DEF_BLOCK_SIZE - Dns.DEF_MAGIC_SIZE)) throw Status.STD_IOERR;
+        Io.out('[DNS] Done. ');
+
+
+        Io.out('[DNS] Writing i-nodes... ');
+
+        var buf = Buf.Buffer();
+        var inode = Dns.Dns_inode(0, [Net.nilcap, Net.nilcap], 0, Afs.Afs_file_state.FF_invalid);
+
+        for (i = 0; i < ninodes; i++) {
+            inode.i_objnum = i;
+            Buf.buf_init(buf);
+            Dns.buf_put_inode(buf, inode);
+            Buf.buf_pad(buf, Dns.DNS_INODE_SIZE);
+            n = Buf.buf_write(part_fd, buf);
+            if (n != Dns.DNS_INODE_SIZE) throw Status.STD_IOERR;
+        }
+        self.dns_super = Dns.Dns_super(label, blocksize, privport, pubport, checkfield, ncols, cols, colmasks,
+                            Dns.Fs_server(fs_caps, [Dns.Fs_state.FS_unknown, Dns.Fs_state.FS_unknown],
+                                          0, Dns.Dns_mode.DNSMODE_ONECOPY));
+        Io.out('[DNS] Done. ');
+        /*
+         ** The live time table. It's a fixed size bitfield table build with
+         ** a string of sufficient size.
+         **
+         ** Assumption: Maximale Live time value < 128, therefore
+         **             7 bit are used for each object. The MSB is
+         **             the used flag (=1 -> inode used).
+         **
+         ** The first entry (obj=0) is used for the lock status
+         ** of the live table.
+         **
+         */
+        Io.out('[DNS] Writing live table... ');
+
+        var live_size = ninodes + 1;
+        var live_table = Buf.Buffer();
+        var live_off = Dns.DEF_BLOCK_SIZE + (ninodes * Dns.DNS_INODE_SIZE);
+
+        function live_set(obj, time, flag) {
+            Buf.buf_set(live_table, obj, ((time & 0x7f) | ((flag << 7) & 0x80)));
+        }
+
+        function live_write() {
+            /*
+             ** Unlock the live table
+             */
+            live_set(0, 0, 1);
+            n = Buf.buf_write(part_fd, live_table, live_off);
+            if (n != live_size) throw Status.STD_IOERR;
+        }
+
+        for (i = 0; i <= ninodes; i++) {
+            live_set(i, Dns.DNS_MAX_LIVE, 0);
+        }
+        Buf.buf_pad(live_table, live_size);
+        live_write();
+
+        Io.out('[DNS] Done. ');
+
+        // w/o cache version, used ony to create the root directory...
+        function write_inode(obj, data, size) {
+            /*
+             ** Read and write full blocks!
+             */
+            var buf = Buf.Buffer();
+            var off = 512 + Dns.off_of_inode(obj);
+            var fileoff = div(off,blocksize)*blocksize;
+            var blockoff = off % blocksize;
+            var n = Buf.buf_read(self.dns_part_fd, buf, fileoff,blocksize);
+            if (n != blocksize)
+                return Status.STD_IOERR;
+            Buf.buf_blit(buf,blockoff,data,0,size);
+            n = Buf.buf_write(self.dns_part_fd, buf, fileoff, blocksize);
+            if (self.sync_write_mode) Io.sync(self.dns_part_fd);
+            if (n == blocksize)
+                return Status.STD_OK;
+            else
+                return Status.STD_IOERR;
+        }
+
+        var rootinode;
+        var rootcap;
+
+        function create_dir(dir,callback) {
+            var stat;
+            var cap = undefined;
+            var obj = dir.dd_objnum;
+            var fs = self.dns_super.dns_fs;
+            var fs_default=fs.fs_default;
+            var buf = Buf.Buffer();
+
+            self.stats.op_create++;
+            Sch.ScheduleBlock([
+                function () {
+                    self.afs.afs_create(fs.fs_cap[fs_default],buf,0,Afs.Afs_commit_flag.AFS_UNCOMMIT,function (_stat,_cap) {
+                        stat=_stat;
+                        cap=_cap;
+                    })
+                },
+                function () {
+                    if (stat!=Status.STD_OK) throw stat;
+                    dir.dd_state=Dns.Dns_dir_state.DD_unlocked;
+                    var fs_caps = [Net.nilcap,Net.nilcap];
+                    fs_caps[fs_default]=cap;
+                    rootinode = Dns.Dns_inode(obj,fs_caps,0,Afs.Afs_file_state.FF_unlocked);
+                    Buf.buf_init(buf);
+                    Dns.buf_put_inode(buf,rootinode);
+                    stat=write_inode(obj,buf,Dns.DNS_INODE_SIZE);
+                    callback(stat);
+                }
+            ],function (e) {
+                if (typeof e == 'number') callback(e); else callback(Status.STD_SYSERR);
+            })
+        }
+
+        function modify_dir(dir,callback) {
+            var stat;
+            var cap = undefined;
+            var obj = dir.dd_objnum;
+            var fs = self.dns_super.dns_fs;
+            var fs_default=fs.fs_default;
+            var dirsize = dir.size();
+            var buf = Buf.Buffer();
+            Dns.buf_put_dir(buf,dir);
+            self.stats.op_create++;
+            Sch.ScheduleBlock([
+                function () {
+                    self.afs.afs_modify(rootinode.i_caps[fs_default],buf,dirsize,0,Afs.Afs_commit_flag.AFS_SAFETY,function (_stat,_cap) {
+                        stat=_stat;
+                        cap=_cap;
+                    })
+                },
+                function () {
+                    if (stat!=Status.STD_OK) throw stat;
+                    dir.dd_state=Dns.Dns_dir_state.DD_locked;
+                    rootinode.i_size=dirsize;
+                    rootinode.i_state=Afs.Afs_file_state.FF_locked;
+                    rootcap=Net.Capability(self.dns_super.dns_putport,
+                                           Net.prv_encode(rootinode.i_objnum,Rights.PRV_ALL_RIGHTS,dir.dd_random));
+                    Buf.buf_init(buf);
+                    Dns.buf_put_inode(buf,rootinode);
+                    stat=write_inode(obj,buf,Dns.DNS_INODE_SIZE);
+                    callback(stat);
+                }
+            ],function (e) {
+                if (typeof e == 'number') callback(e); else callback(Status.STD_SYSERR);
+            })
+        }
+
+        /** Create the root directory (Object number 1) and return it and the status
+         ** of the server create function. All pending writes are already done here.
+         **
+         ** Note: the root directory can't be deleted. Therefore, it's not
+         ** garbage collected.
+         *
+         */
+        function create_root(callback) {
+            var stat;
+            var sup=self.dns_super;
+            self.stats.op_create++;
+            assert(sup.try_lock());
+            if (sup.dns_nextfree!=1) {
+                sup.unlock();
+                callback(Status.STD_EXISTS);
+                return;
+            }
+            sup.dns_nextfree++;
+            var rootdir = Dns.Dns_dir(1,sup.dns_ncols,0,sup.dns_colnames,Net.uniqport(),[],
+                                      Dns.Dns_dir_state.DD_unlocked,self.time(),Dns.DNS_MAX_LIVE);
+            rootdir.dd_lock.acquire(); // Non-blocking, we're the first one
+            sup.unlock();
+            Sch.ScheduleBlock([
+                function () {
+                    create_dir(rootdir,function (_stat) {stat=_stat});
+                },
+                function () {
+                    if (stat!=Status.STD_OK) {
+                        throw stat;
+                    }
+                    modify_dir(rootdir,function (_stat) {
+                        stat=_stat;
+                    });
+                }, function () {
+                    rootdir.dd_lock.release();
+                    callback(stat,rootdir);
+                }
+            ], function (e) {
+                if (rootdir!=undefined) rootdir.dd_lock.release();
+                if (typeof e == 'number') callback(e,undefined); else callback(Status.STD_SYSERR,undefined);
+            });
+        }
+
+
+        var dir = undefined;
+        Sch.ScheduleBlock([
+            function () {
+                Io.out('[DNS] Creating the root directory ... ');
+                create_root(function (_stat,_dir) {
+                    stat = _stat;
+                    dir = _dir;
+                })
+            },
+            function () {
+                if (stat == Status.STD_OK) {
+                    Io.out('[DNS] Done.');
+                    self.rootdir = dir;
+                }
+                else Io.out('[DNS] Creation of root directory failed: ' + Status.print(stat));
+                Io.close(part_fd);
+                self.dns_part_fd=undefined;
+                Io.out('[DNS] Finished.');
+                callback(stat);
+            }
+        ])
+    }
+    try {
+        create (callback);
+    } catch (e) {
+        if (self.dns_part_fd) Io.close(self.dns_part_fd);
+        self.dns_part_fd=undefined;
+        if (typeof e == 'number') callback(e); else {
+            Io.printstack(e,'Dns_server.create_fs');
+            callback(Status.STD_IOERR);
+        }
+        throw Error(e);
+    }
+};
+
+/** Open the DNS system (one inode partition, UNIX file) [BLOCKING]
+ **
+ **  part_inode:     Directory Mapping Partition File
+ **  cache:      Cache parameters
+ *
+ *
+ * @param {string} part_inode
+ * @param {dns_cache_parameter} cache_params
+ * @param {function(Status.STD_OK|*)} callback
+ */
+dns_server.prototype.open_fs = function (part_inode,cache_params,callback) {
+    /**
+     *
+     * @type {dns_server}
+     */
+    var self = this;
+    self.dns_super=Dns.Dns_super();
+    var sup;
+
+    function open(callback) {
+        var n, i,j;
+        self.dns_super=Dns.Dns_super();
+        sup = self.dns_super;
+
+        var stat = Status.STD_OK;
+        Io.out('[DNS] Opening DNS ...');
+        if (Io.exists(part_inode)) {
+            Io.out('[DNS] Using directory i-node partition ' + part_inode);
+        } else {
+            Io.out('[DNS] No existing directory i-node partition ' + part_inode);
+            throw Status.STD_IOERR;
+        }
+        self.dns_part = part_inode;
+        self.dns_part_fd = Io.open(part_inode, (self.iocache_bypass ? 'rs+' : 'r+'));
+        /*
+         ** Either the filesystem is read only or we have an
+         ** inconsistent filesystem. In that case, only read
+         ** request are allowed.
+         */
+        var readonly = false;
+        Io.out('[DNS] Reading the super block...');
+        var superbuf = Buf.Buffer();
+        n = Buf.buf_read(self.dns_part_fd, superbuf, 0, Dns.DEF_BLOCK_SIZE);
+        if (n != Dns.DEF_BLOCK_SIZE) {
+            Io.out('[DNS] Cannot read super block.');
+            Io.close(self.dns_part_fd);
+            throw Status.STD_IOERR;
+        }
+        var magic1 = Buf.buf_get_string(superbuf);
+        if (!String.equal(magic1, Dns.MAGIC_STR)) {
+            Io.out('[DNS] Invalid i-node partition magic found, got ' + magic1 + ', but exptected ' + Dns.MAGIC_STR);
+            throw Status.STD_IOERR;
+        } else {
+            Io.out('[DNS] Found i-node partition magic  ' + magic1);
+        }
+        Buf.buf_set_pos(superbuf, Dns.DEF_MAGIC_SIZE);
+        sup.dns_name = Buf.buf_get_string(superbuf);
+        Buf.buf_set_pos(superbuf, 256 + Dns.DEF_MAGIC_SIZE);
+        sup.dns_ndirs = Buf.buf_get_int32(superbuf);
+        sup.dns_block_size = Buf.buf_get_int32(superbuf);
+        sup.dns_getport = Buf.buf_get_port(superbuf);
+        sup.dns_putport = Buf.buf_get_port(superbuf);
+        sup.dns_checkfield = Buf.buf_get_port(superbuf);
+        sup.dns_ncols = Buf.buf_get_int32(superbuf);
+        var cols=[];
+        var colmasks=[];
+
+        for(i=0;i<sup.dns_ncols;i++) {
+            cols.push(Buf.buf_get_string(superbuf));
+        }
+        for(i=0;i<sup.dns_ncols;i++) {
+            colmasks.push(DnsInt.buf_get_rights(superbuf));
+        }
+        var fs_caps=[];
+        for(i=0;i<2;i++) {
+            var cap=Buf.buf_get_cap(superbuf);
+            if (!Net.Equal.capability(cap,Net.nilcap)) Io.out('[DNS] Using AFS  ' + Net.Print.capability(cap));
+            fs_caps.push(cap);
+        }
+
+        sup.dns_colnames=cols;
+        sup.dns_generic_colmask=colmasks;
+        sup.dns_fs=Dns.Fs_server(fs_caps, [Dns.Fs_state.FS_unknown, Dns.Fs_state.FS_unknown],
+            0, Dns.Dns_mode.DNSMODE_ONECOPY);
+        /*
+         ** Defaulf file server and supported mode must be located later (get_def_fs)
+         */
+        Io.out('  DNS: Label: '+sup.dns_name);
+        Io.out('  DNS: Maximal number of directories: '+sup.dns_ndirs);
+        Io.out('  DNS: Blocksize: '+sup.dns_block_size);
+        /*
+         ** The livtime table. It's a fixed size bitfield table build with
+         ** a string of sufficient size.
+         **
+         ** Assumption: Maximale Live time value < 128, therefore
+         **             7 bit are used for each object. The MSB is
+         **             the used flag (=1 -> inode used).
+         **
+         ** The first entry (obj=0) is used for the lock status
+         ** of the live table.
+         */
+        var live_size = sup.dns_ndirs+1;
+        var live_table = Buf.Buffer();
+        var live_off = 512 + (sup.dns_ndirs*Dns.DNS_INODE_SIZE)
+        /**
+         *
+         * @param {number} obj
+         * @param {number} time
+         * @param {number} flag
+         */
+        function live_set(obj, time, flag) {
+            Buf.buf_set(live_table, obj, ((time & 0x7f) | ((flag << 7) & 0x80)));
+        }
+
+        /**
+         *
+         * @param {number} obj
+         * @returns {{time: number, flag: number}}
+         */
+        function live_get(obj) {
+            var v = Buf.buf_get(live_table, obj);
+            return {time: (v & 0x7f), flag: ((v & 0x80) >> 7)}
+        }
+
+        /**
+         *
+         * @returns {number}
+         */
+        function live_read() {
+            var n = Buf.buf_read(self.dns_part_fd, live_table, live_off, live_size);
+            if (n != live_size) {
+                return Status.STD_IOERR;
+            }
+            var live = live_get(0);
+            if (live.time == 0 && live.flag == 1) {
+                /*
+                 ** Now lock the live table. After a server crash,
+                 ** the restarted server will found the lock and must
+                 ** discard the live table. All server objects got
+                 ** the maximal live time!
+                 */
+                live_set(0, 1, 1);
+                n = Buf.buf_write(self.dns_part_fd, live_table, live_off, 1);
+                if (n != 1) return Status.STD_IOERR;
+                else return Status.STD_OK;
+            } else return Status.STD_ARGBAD;
+        }
+
+        /**
+         *
+         * @returns {number}
+         */
+        function live_write() {
+            /*
+             ** Unlock the live table
+             */
+            live_set(0, 0, 1);
+            n = Buf.buf_write(self.dns_part_fd, live_table, live_off);
+            if (n != live_size) return Status.STD_IOERR;
+            if (self.sync_write_mode) Io.sync(self.dns_part_fd);
+            return Status.STD_OK;
+        }
+
+        self.live_set = live_set;
+        self.live_get = live_get;
+        self.live_read = live_read;
+        self.live_write = live_write;
+
+        /*
+         ** Read the live table
+         */
+        Io.out('[DNS] Reading the live time table...');
+        stat = live_read();
+        if (stat == Status.STD_ARGBAD) {
+            Io.out('[DNS] Found locked live table: Reinitialize...');
+            for (i = 0; i < sup.dns_ndirs - 1; i++) {
+                live_set(i, Dns.DNS_MAX_LIVE, 0);
+            }
+        } else if (stat == Status.STD_IOERR) {
+            Io.out('[DNS] IO Error.');
+            Io.close(self.dns_part_fd);
+            throw Status.STD_IOERR;
+        }
+        Io.out('[DNS] Ok.');
+        stat=Status.STD_OK;
+
+        /**
+         ** Raw disk Read and Write functions for the cache module.
+         **
+         ** Units:
+         **  addr: blocks
+         **  size: bytes
+         **
+         */
+        /**
+         *
+         * @param obj
+         * @param addr
+         * @param data
+         * @param size
+         * @returns {number}
+         */
+        self.read_inode = function (obj, addr, data, size) {
+            var off = addr * sup.dns_block_size;
+            Io.log((log < 10) || ('read_inode obj=' + obj + ' addr=' + addr + ' size=' + size));
+            var n = Buf.buf_read(self.dns_part_fd, data, off, size);
+            if (n == size)
+                return Status.STD_OK;
+            else
+                return Status.STD_SYSERR;
+
+        };
+
+        /**
+         *
+         * @param obj
+         * @param addr
+         * @param data
+         * @param size
+         * @returns {number}
+         */
+        self.write_inode = function (obj, addr, data, size) {
+            var off = addr * sup.dns_block_size;
+            Io.log((log < 10) || ('write_inode obj=' + obj + ' addr=' + addr + ' size=' + size));
+            var n = Buf.buf_write(self.dns_part_fd, data, off, size);
+            if (self.sync_write_mode) Io.sync(self.dns_part_fd);
+            if (n == size)
+                return Status.STD_OK;
+            else
+                return Status.STD_SYSERR;
+        };
+
+        self.sync_inode = function (obj) {
+        };
+
+        Io.out('[DNS] Creating inode cache ' + cache_params.c_inode_size + '[' + cache_params.c_inode_buffers + '] ... ');
+
+        /*
+         ** Note: I-nodes are handled only on buffer block level
+         ** #i-node=cache_params.c_inode_size/DEF_INODE_SIZE
+         */
+        var cache_inode =
+            Fcache.Fcache(
+                'DNS Inode',
+                cache_params.c_inode_buffers,
+                Dns.DEF_BLOCK_SIZE,
+                cache_params.c_inode_size,
+                self.read_inode,
+                self.write_inode,
+                self.sync_inode,
+                Fcache.Fs_cache_mode.Cache_R);
+        self.cache_inode = cache_inode;
+        if (self.cache_inode == undefined) {
+            Io.out('[DNS] IO Error.');
+            Io.close(self.dns_part_fd);
+            throw Status.STD_IOERR;
+        }
+        /*
+         ** The i-node cache object. One object for all i-nodes!
+         ** Disk address = obj = 1 [blocks]
+         */
+        var res = cache_inode.cache_lookup(1, 1, Dns.off_of_inode(sup.dns_ndirs), Afs.Afs_file_state.FF_unlocked);
+        var inode_fse = res.fse;
+        self.inode_cache_entry = inode_fse;
+
+        /*
+         ** Some i-node utils
+         */
+        self.inode_of_obj = function(obj) {
+            try {
+                var inode = undefined;
+                var buf = Buf.Buffer();
+                var stat = cache_inode.cache_read(inode_fse, buf, Dns.off_of_inode(obj), Dns.DNS_INODE_SIZE);
+                if (stat != Status.STD_OK) return {stat: stat, inode: undefined};
+                inode = Dns.buf_get_inode(buf);
+                return {stat: stat, inode: inode};
+            } catch (e) {
+                return {stat: Status.STD_SYSERR, inode: undefined};
+            }
+        };
+
+        /**
+         *
+         * @param {number} obj
+         * @returns {{stat:(Status.STD_OK|*),cap:(capability|undefined),inode:(dns_inode|undefined)}}
+         */
+         self.cap_of_obj = function(obj) {
+            var stat;
+            var si = self.inode_of_obj(obj);
+            if (si.stat != Status.STD_OK) return {stat:stat,cap:undefined,inode:undefined};
+            var cap = si.inode.i_caps[self.fs_default];
+            if (si.inode.i_state == Afs.Afs_file_state.FF_locked ||
+                si.inode.i_state == Afs.Afs_file_state.FF_unlocked) stat=Status.STD_OK;
+            else stat=Status.STD_SYSERR;
+            return {stat:stat,cap:cap,inode:si.inode};
+        };
+
+        /**
+         *
+         * @param {dns_inode} inode
+         * @returns {Status.STD_OK|*}
+         */
+         self.update_inode = function(inode) {
+            try {
+                var buf = Buf.Buffer();
+                Dns.buf_put_inode(buf, inode);
+                var stat = cache_inode.cache_write(inode_fse, buf, Dns.off_of_inode(inode.i_objnum),Dns.DEF_INODE_SIZE);
+                return stat;
+            } catch (e) {
+                return Status.STD_SYSERR;
+            }
+        };
+
+        var cache_data = Cache.Cache(cache_params.c_dir_buffers);
+
+        self.cache_data = cache_data;
+
+        if (cache_data == undefined) {
+            Io.out('[DNS] IO Error.');
+            Io.close(self.dns_part_fd);
+            throw Status.STD_IOERR;
+        }
+
+        /*
+         ** Round up the value x to block_size
+         */
+
+        function block(x) {
+            return Dns.ceil_block_bytes(x, sup.dns_block_size);
+        }
+
+        function to_block(x) {
+            return div(x, sup.dns_block_size);
+        }
+
+        function of_block(x) {
+            return (x * sup.dns_block_size);
+        }
+
+        self.block = block;
+        self.to_block = to_block;
+        self.of_block = of_block;
+        /*
+         ** Read the i-node table.
+         ** Build a list with free i-node object numbers below the
+         ** nextfree boundary (above nextfree there are all
+         ** free i-nodes, if any).
+         */
+        Io.out('[DNS] reading the i-node table ...');
+        var freeino=[];
+        var firstfree=-1;
+        var nextfree=-1;
+        var nused=0;
+        var inodeb = Buf.Buffer();
+        for(i=1;i<sup.dns_ndirs;i++) {
+            stat = cache_inode.cache_read(inode_fse, inodeb, Dns.off_of_inode(i), Dns.DEF_INODE_SIZE);
+            if (stat != Status.STD_OK) {
+                Io.out('[DNS] Failed to read i-node ' + i + ': ' + Status.print(stat));
+                throw stat;
+            }
+            inodeb.pos = 0;
+            var inode = Dns.buf_get_inode(inodeb);
+            /*
+             ** Some sanity checks first
+             */
+
+            if (inode.i_objnum != i) {
+                Fs.closeSync(self.dns_part_fd);
+                Io.out('[DNS] got invalid i-node ' + i + ': ' + util.inspect(inode));
+                Io.out('[DNS] Abort.');
+                throw Status.STD_SYSERR;
+            }
+            if (inode.i_state == Afs_file_state.FF_invalid ||
+                inode.i_state == Afs_file_state.FF_unlocked) {
+
+                live_set(i, 0, 0);
+
+                if (nextfree != -1 && nextfree != (i - 1)) {
+
+                    for (j = firstfree; j <= nextfree; j++) {
+                        freeino.push(j);
+                    }
+                    firstfree = i;
+                    nextfree = i;
+                } else {
+                    nextfree = i;
+                    if (firstfree == -1)
+                        firstfree = i;
+                }
+
+                if (inode.i_state == Afs_file_state.FF_unlocked) {
+                    Io.out('[DNS] Unlocked/uncommitted directory found. Destroy it: ' + inode.print());
+                    inode.i_state = Afs_file_state.FF_invalid;
+                    inodeb.pos = 0;
+                    Dns.buf_put_inode(inodeb, inode);
+                    stat = cache_inode.cache_write(inode_fse, inodeb, Dns.off_of_inode(i), Dns.DEF_INODE_SIZE);
+
+                    if (stat != Status.STD_OK)
+                        throw stat;
+
+                }
+            } else {
+                var tf = live_get(i);
+                live_set(i, tf.time, 1);
+                nused++;
+            }
+        }
+        Io.out('[DNS] Ok.');
+
+        if (nextfree != -1 && nextfree < (sup.dns_ndirs - 1)) {
+
+            /*
+             ** There are only used i-nodes at the end of the i-node table.
+             */
+            for (j = firstfree; j <= nextfree; j++) {
+                freeino.push(j);
+            }
+            nextfree = -1;
+        } else if (firstfree == 0 && nextfree == (sup.dns_ndirs - 1)) {
+
+            /*
+             ** Empty filesystem.
+             */
+            nextfree = 1;
+        } else if (firstfree > 0) {
+
+            /*
+             ** There are free i-nodes at the end of the i-node table.
+             */
+            nextfree = firstfree;
+        }
+
+
+        sup.dns_nused = nused;
+        sup.dns_freeobjnums = freeino;
+        sup.dns_nextfree = nextfree;
+
+        Io.out('[DNS] Found ' + sup.dns_nused + ' used Inode(s)');
+
+        Io.log((log < 0) || (self.dns_super.print()));
+        /*
+        ** Read the root directory...
+         */
+        self.acquire_dir(1,function (_stat,_dir) {
+            stat=_stat;
+            self.rootdir=_dir;
+            callback(stat);
+        });
+
+    }
+    try {
+        open(callback);
+    } catch (e) {
+        if (typeof e == 'number') callback(e); else {
+            Io.printstack(e,'Dns_server.open_fs');
+            callback(Status.STD_IOERR);
+        }
+        // throw Error(e);
+    }
+};
+
+
+/*
+ ** ================
+ ** PUBLIC INTERFACE
+ ** ================
+ */
+
+var DnsRpc = Require('dos/dns_srv_rpc');
+//afs_server.prototype=new AfsRpc.afs_server_rpc();
+dns_server.prototype.dns_lookup=DnsRpc.dns_server_rpc.prototype.dns_lookup;
+dns_server.prototype.dns_create=DnsRpc.dns_server_rpc.prototype.dns_create;
+dns_server.prototype.dns_delete=DnsRpc.dns_server_rpc.prototype.dns_delete;
+dns_server.prototype.dns_delete_dir=DnsRpc.dns_server_rpc.prototype.dns_delete_dir;
+dns_server.prototype.dns_append=DnsRpc.dns_server_rpc.prototype.dns_append;
+dns_server.prototype.dns_rename=DnsRpc.dns_server_rpc.prototype.dns_rename;
+dns_server.prototype.dns_replace=DnsRpc.dns_server_rpc.prototype.dns_replace;
+dns_server.prototype.dns_info=DnsRpc.dns_server_rpc.prototype.dns_info;
+dns_server.prototype.dns_stat=DnsRpc.dns_server_rpc.prototype.dns_stat;
+dns_server.prototype.dns_touch=DnsRpc.dns_server_rpc.prototype.dns_touch;
+dns_server.prototype.dns_age=DnsRpc.dns_server_rpc.prototype.dns_age;
+dns_server.prototype.dns_restrict=DnsRpc.dns_server_rpc.prototype.dns_restrict;
+dns_server.prototype.dns_exit=DnsRpc.dns_server_rpc.prototype.dns_exit;
+dns_server.prototype.dns_list=DnsRpc.dns_server_rpc.prototype.dns_list;
+dns_server.prototype.dns_setlookup=DnsRpc.dns_server_rpc.prototype.dns_setlookup;
+dns_server.prototype.dns_start_server=DnsRpc.dns_server_rpc.prototype.dns_start_server;
+
+module.exports = {
+    /**
+     *
+     * @param rpc
+     * @returns {dns_server}
+     */
+    Server: function(rpc,options) {
+        var obj = new dns_server(rpc,options);
+        Object.preventExtensions(obj);
+        return obj;
+    }
+};