Search Results

Search found 10789 results on 432 pages for 'cpu upgrade'.

Page 209/432 | < Previous Page | 205 206 207 208 209 210 211 212 213 214 215 216  | Next Page >

• Does the video card influence the speed at which a video is processed and exported from Adobe Premie

  - by Xv

  or is all the processing done by the CPU. What is an entry level card that would handle video processing adequately?

  Read the article

• Message from Nagios Server

  - by user12213

  Nagios Server is monitoring my Server which hosts Windows Sharepoint. I am getting the following 2 alerts in my inbox from Nagios Server 1. Service: C:\ Drive Space State: CRITICAL Additional Info: CRITICAL - Socket timeout after 10 seconds 2. Service: CPU Load State: CRITICAL Additional Info: CRITICAL - Socket timeout after 10 seconds What do I infer from these?

  Read the article

• Using Google Docs as a Storage System like S3. Is there any limit?

  - by mickthomp

  Hi all, I'm considering to upgrade to a Google Docs Premium Account (gDrive)? I'm wondering if that can be used as I'm using Amazon S3 at the moment. I'd like to upload images. Do you know if there is there any limit on the number of images I can upload on my 200GB Google Docs account? I think it could be really useful to have something like that and we could save some money on our webapps. Thank you ;)

  Read the article

• 100% utilization on amazon server

  - by user2939830

  Good day, I would just like to know if you guys have any idea what could be the possible cause for a sudden disconnection of clients and 100% cpu utilization in our amazon server. This problem just started 2 days ago and in both occasion it happened at 7 plus in the morning gmt+8. What we usually do is just reset the socket for it to normalize and then on the next day same thing happened at 7 in the morning every client is disconnected from the server.

  Read the article

• Where are stickies ( Sticky Notes) stored on mac 10.9.3?

  - by user332203

  i deleted an important note on stickies. And i retrieved an old version of it in time machine under preferences / widgets. but the setup appears to have changed in my upgrade to mavericks and I can't open the note. I'm trying to open a "post-mavericks" version in my time machine and I can't find where it is. i saw a post that said look under Library/Preferences/Container, i have no such folder or binary document. Please help.

  Read the article

• my gateway laptop will not reboot

  - by dom

  My gateway laptop model nx570xl originally rebooted normally. I don't know when it happened, but now when i try to reboot it Keeps trying but it never happens forcing me to shut down and wait a random amt of time before it will start up again. Its very annoying and wastes a lot of my time. I don't think its a cpu overheating problem because its random when it starts up after i turn it back on. Any ideas?

  Read the article

• Problem with room/screen/menu controller in python game: old rooms are not removed from memory

  - by Jordan Magnuson

  I'm literally banging my head against a wall here (as in, yes, physically, at my current location, I am damaging my cranium). Basically, I've got a Python/Pygame game with some typical game "rooms", or "screens." EG title screen, high scores screen, and the actual game room. Something bad is happening when I switch between rooms: the old room (and its various items) are not removed from memory, or from my event listener. Not only that, but every time I go back to a certain room, my number of event listeners increases, as well as the RAM being consumed! (So if I go back and forth between the title screen and the "game room", for instance, the number of event listeners and the memory usage just keep going up and up. The main issue is that all the event listeners start to add up and really drain the CPU. I'm new to Python, and don't know if I'm doing something obviously wrong here, or what. I will love you so much if you can help me with this! Below is the relevant source code. Complete source code at http://www.necessarygames.com/my_games/betraveled/betraveled_src0328.zip MAIN.PY class RoomController(object): """Controls which room is currently active (eg Title Screen)""" def __init__(self, screen, ev_manager): self.room = None self.screen = screen self.ev_manager = ev_manager self.ev_manager.register_listener(self) self.room = self.set_room(config.room) def set_room(self, room_const): #Unregister old room from ev_manager if self.room: self.room.ev_manager.unregister_listener(self.room) self.room = None #Set new room based on const if room_const == config.TITLE_SCREEN: return rooms.TitleScreen(self.screen, self.ev_manager) elif room_const == config.GAME_MODE_ROOM: return rooms.GameModeRoom(self.screen, self.ev_manager) elif room_const == config.GAME_ROOM: return rooms.GameRoom(self.screen, self.ev_manager) elif room_const == config.HIGH_SCORES_ROOM: return rooms.HighScoresRoom(self.screen, self.ev_manager) def notify(self, event): if isinstance(event, ChangeRoomRequest): if event.game_mode: config.game_mode = event.game_mode self.room = self.set_room(event.new_room) #Run game def main(): pygame.init() screen = pygame.display.set_mode(config.screen_size) ev_manager = EventManager() spinner = CPUSpinnerController(ev_manager) room_controller = RoomController(screen, ev_manager) pygame_event_controller = PyGameEventController(ev_manager) spinner.run() EVENT_MANAGER.PY class EventManager: #This object is responsible for coordinating most communication #between the Model, View, and Controller. def __init__(self): from weakref import WeakKeyDictionary self.last_listeners = {} self.listeners = WeakKeyDictionary() self.eventQueue= [] self.gui_app = None #---------------------------------------------------------------------- def register_listener(self, listener): self.listeners[listener] = 1 #---------------------------------------------------------------------- def unregister_listener(self, listener): if listener in self.listeners: del self.listeners[listener] #---------------------------------------------------------------------- def clear(self): del self.listeners[:] #---------------------------------------------------------------------- def post(self, event): # if isinstance(event, MouseButtonLeftEvent): # debug(event.name) #NOTE: copying the list like this before iterating over it, EVERY tick, is highly inefficient, #but currently has to be done because of how new listeners are added to the queue while it is running #(eg when popping cards from a deck). Should be changed. See: http://dr0id.homepage.bluewin.ch/pygame_tutorial08.html #and search for "Watch the iteration" print 'Number of listeners: ' + str(len(self.listeners)) for listener in list(self.listeners): #NOTE: If the weakref has died, it will be #automatically removed, so we don't have #to worry about it. listener.notify(event) def notify(self, event): pass #------------------------------------------------------------------------------ class PyGameEventController: """...""" def __init__(self, ev_manager): self.ev_manager = ev_manager self.ev_manager.register_listener(self) self.input_freeze = False #---------------------------------------------------------------------- def notify(self, incoming_event): if isinstance(incoming_event, UserInputFreeze): self.input_freeze = True elif isinstance(incoming_event, UserInputUnFreeze): self.input_freeze = False elif isinstance(incoming_event, TickEvent) or isinstance(incoming_event, BoardCreationTick): #Share some time with other processes, so we don't hog the cpu pygame.time.wait(5) #Handle Pygame Events for event in pygame.event.get(): #If this event manager has an associated PGU GUI app, notify it of the event if self.ev_manager.gui_app: self.ev_manager.gui_app.event(event) #Standard event handling for everything else ev = None if event.type == QUIT: ev = QuitEvent() elif event.type == pygame.MOUSEBUTTONDOWN and not self.input_freeze: if event.button == 1: #Button 1 pos = pygame.mouse.get_pos() ev = MouseButtonLeftEvent(pos) elif event.type == pygame.MOUSEBUTTONDOWN and not self.input_freeze: if event.button == 2: #Button 2 pos = pygame.mouse.get_pos() ev = MouseButtonRightEvent(pos) elif event.type == pygame.MOUSEBUTTONUP and not self.input_freeze: if event.button == 2: #Button 2 Release pos = pygame.mouse.get_pos() ev = MouseButtonRightReleaseEvent(pos) elif event.type == pygame.MOUSEMOTION: pos = pygame.mouse.get_pos() ev = MouseMoveEvent(pos) #Post event to event manager if ev: self.ev_manager.post(ev) # elif isinstance(event, BoardCreationTick): # #Share some time with other processes, so we don't hog the cpu # pygame.time.wait(5) # # #If this event manager has an associated PGU GUI app, notify it of the event # if self.ev_manager.gui_app: # self.ev_manager.gui_app.event(event) #------------------------------------------------------------------------------ class CPUSpinnerController: def __init__(self, ev_manager): self.ev_manager = ev_manager self.ev_manager.register_listener(self) self.clock = pygame.time.Clock() self.cumu_time = 0 self.keep_going = True #---------------------------------------------------------------------- def run(self): if not self.keep_going: raise Exception('dead spinner') while self.keep_going: time_passed = self.clock.tick() fps = self.clock.get_fps() self.cumu_time += time_passed self.ev_manager.post(TickEvent(time_passed, fps)) if self.cumu_time >= 1000: self.cumu_time = 0 self.ev_manager.post(SecondEvent(fps=fps)) pygame.quit() #---------------------------------------------------------------------- def notify(self, event): if isinstance(event, QuitEvent): #this will stop the while loop from running self.keep_going = False EXAMPLE CLASS USING EVENT MANAGER class Timer(object): def __init__(self, ev_manager, time_left): self.ev_manager = ev_manager self.ev_manager.register_listener(self) self.time_left = time_left self.paused = False def __repr__(self): return str(self.time_left) def pause(self): self.paused = True def unpause(self): self.paused = False def notify(self, event): #Pause Event if isinstance(event, Pause): self.pause() #Unpause Event elif isinstance(event, Unpause): self.unpause() #Second Event elif isinstance(event, SecondEvent): if not self.paused: self.time_left -= 1

  Read the article

• Updating resources in SharpDX - why can I not map a dynamic texture?

  - by sebf

  I am trying to map a Texture2D resource in DirectX11 via SharpDX. The resource is declared as a ShaderResource, with Default usage and the 'Write' CPU flag specified. My call however fails with a generic exception from SharpDX: _Parent.Context.MapSubresource(_Resource, 0, SharpDX.Direct3D11.MapMode.Write, SharpDX.Direct3D11.MapFlags.None, out stream); I see from this question that it is supported. The MSDN docs and this other question hint that instead of using Context.MapSubresource() I should be using Texture2D.Map(), however, the DirectX11 Texture2D class does not define Map() (though it does for the DX10 equivalent). If I call the above with MapMode.WriteDiscard, the call succeeds but in this case the previous content of the texture is lost, which is no good when I only want to update a section of it. Has the Map() method been removed in DirectX11 or am I looking in the wrong place? Is the MapSubresource() method unsuitable or am I using it wrong?

  Read the article

• dropbox slow on ubuntu 14.04

  - by Donbeo

  My dropbox syncing is incredibly slow... I am using dropbox from the ubuntu repository on an almost fresh ubuntu installation. I would like to avoid to install the package from the dropbox website for the reasons explained here Dropbox Upgrade Is someone having the same problem? How can I solve? EDIT : This is an example of what I get. donbeo@donbeo-HP-EliteBook-Folio-9470m:~$ dropbox status Syncing (17 files remaining, 22 secs left) Uploading 17 files (123.3 KB/sec, 22 secs left) donbeo@donbeo-HP-EliteBook-Folio-9470m:~$ dropbox status Syncing (17 files remaining, 3 mins left) Uploading 17 files (13.2 KB/sec, 3 mins left) donbeo@donbeo-HP-EliteBook-Folio-9470m:~$ dropbox status Syncing (17 files remaining, 5 mins left) Uploading 17 files (8.2 KB/sec, 5 mins left) donbeo@donbeo-HP-EliteBook-Folio-9470m:~$ dropbox status Syncing (17 files remaining) Uploading 17 files... donbeo@donbeo-HP-EliteBook-Folio-9470m:~$ dropbox status Syncing (17 files remaining) Uploading 17 files... donbeo@donbeo-HP-EliteBook-Folio-9470m:~$ EDIT: I have run sudo dropbox update so I am probably using the last version of dropbox

  Read the article

• Cumulative Update packages for SQL Server 2008 R2 RTM & SQL Server 2008 SP1

  - by ssqa.net

  Here is the news on Cumulative Update release news on SQL Server 2008 R2 RTM & SQL Server 2008 Service Pack 1. First let us go through SQL Server 2008 R2 RTM cumulative update, release consist the only hotfixes that were released in Cumulative Update 5, 6, & 7 for SQL Server 2008 SP1. Cumulative Update 1 for SQL 2008 R2 RTM is only intended as a post-RTM rollup for Cumulative Update 5-7 for the release version of SQL Server 2008 SP1 customers who plan to upgrade to SQL Server 2008 R2 and...(read more)

  Read the article

• LWJGL - Eclipse error [on hold]

  - by Zarkopafilis

  When I try to run my lwjgl project, an error pops . Here is the log file: # A fatal error has been detected by the Java Runtime Environment: # EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0x6d8fcc0a, pid=5612, tid=900 # JRE version: 6.0_16-b01 Java VM: Java HotSpot(TM) Client VM (14.2-b01 mixed mode windows-x86 ) Problematic frame: V [jvm.dll+0xfcc0a] # If you would like to submit a bug report, please visit: http://java.sun.com/webapps/bugreport/crash.jsp # --------------- T H R E A D --------------- Current thread (0x016b9000): JavaThread "main" [_thread_in_vm, id=900, stack(0x00160000,0x001b0000)] siginfo: ExceptionCode=0xc0000005, reading address 0x00000000 Registers: EAX=0x00000000, EBX=0x00000000, ECX=0x00000006, EDX=0x00000000 ESP=0x001af4d4, EBP=0x001af524, ESI=0x016b9000, EDI=0x016b9110 EIP=0x6d8fcc0a, EFLAGS=0x00010246 Top of Stack: (sp=0x001af4d4) 0x001af4d4: 6da44bd8 016b9110 00000000 001af668 0x001af4e4: ffffffff 22200000 001af620 76ec39c2 0x001af4f4: 001af524 6d801086 0000000b 001afd34 0x001af504: 016b9000 016dd990 016b9000 00000000 0x001af514: 001af5f4 6d9ee000 6d9ef2f0 ffffffff 0x001af524: 001af58c 10008c85 016b9110 00000000 0x001af534: 00000000 000a0554 00000000 00000024 0x001af544: 00000000 00000000 001af6ac 00000000 Instructions: (pc=0x6d8fcc0a) 0x6d8fcbfa: e8 e8 d0 1d 08 00 8b 45 10 c7 45 d8 0b 00 00 00 0x6d8fcc0a: 8b 00 8b 48 08 0f b7 51 26 8b 40 0c 8b 4c 90 20 Stack: [0x00160000,0x001b0000], sp=0x001af4d4, free space=317k Native frames: (J=compiled Java code, j=interpreted, Vv=VM code, C=native code) V [jvm.dll+0xfcc0a] C [lwjgl.dll+0x8c85] C [USER32.dll+0x18876] C [USER32.dll+0x170f4] C [USER32.dll+0x1119e] C [ntdll.dll+0x460ce] C [USER32.dll+0x10e29] C [USER32.dll+0x10e84] C [lwjgl.dll+0x1cf0] j org.lwjgl.opengl.WindowsDisplay.createWindow(Lorg/lwjgl/opengl/DrawableLWJGL;Lorg/lwjgl/opengl/DisplayMode;Ljava/awt/Canvas;II)V+102 j org.lwjgl.opengl.Display.createWindow()V+71 j org.lwjgl.opengl.Display.create(Lorg/lwjgl/opengl/PixelFormat;Lorg/lwjgl/opengl/Drawable;Lorg/lwjgl/opengl/ContextAttribs;)V+72 j org.lwjgl.opengl.Display.create(Lorg/lwjgl/opengl/PixelFormat;)V+12 j org.lwjgl.opengl.Display.create()V+7 j zarkopafilis.koding.io.javafx.Main.main([Ljava/lang/String;)V+16 v ~StubRoutines::call_stub V [jvm.dll+0xecf9c] V [jvm.dll+0x1741e1] V [jvm.dll+0xed01d] V [jvm.dll+0xf5be5] V [jvm.dll+0xfd83d] C [javaw.exe+0x2155] C [javaw.exe+0x833e] C [kernel32.dll+0x51154] C [ntdll.dll+0x5b2b9] C [ntdll.dll+0x5b28c] Java frames: (J=compiled Java code, j=interpreted, Vv=VM code) j org.lwjgl.opengl.WindowsDisplay.nCreateWindow(IIIIZZJ)J+0 j org.lwjgl.opengl.WindowsDisplay.createWindow(Lorg/lwjgl/opengl/DrawableLWJGL;Lorg/lwjgl/opengl/DisplayMode;Ljava/awt/Canvas;II)V+102 j org.lwjgl.opengl.Display.createWindow()V+71 j org.lwjgl.opengl.Display.create(Lorg/lwjgl/opengl/PixelFormat;Lorg/lwjgl/opengl/Drawable;Lorg/lwjgl/opengl/ContextAttribs;)V+72 j org.lwjgl.opengl.Display.create(Lorg/lwjgl/opengl/PixelFormat;)V+12 j org.lwjgl.opengl.Display.create()V+7 j zarkopafilis.koding.io.javafx.Main.main([Ljava/lang/String;)V+16 v ~StubRoutines::call_stub --------------- P R O C E S S --------------- Java Threads: ( = current thread ) 0x0179a400 JavaThread "Low Memory Detector" daemon [_thread_blocked, id=4460, stack(0x0b900000,0x0b950000)] 0x01795400 JavaThread "CompilerThread0" daemon [_thread_blocked, id=5264, stack(0x0b8b0000,0x0b900000)] 0x01790c00 JavaThread "Attach Listener" daemon [_thread_blocked, id=6080, stack(0x0b860000,0x0b8b0000)] 0x01786400 JavaThread "Signal Dispatcher" daemon [_thread_blocked, id=1204, stack(0x0b810000,0x0b860000)] 0x01759c00 JavaThread "Finalizer" daemon [_thread_blocked, id=5772, stack(0x0b7c0000,0x0b810000)] 0x01755000 JavaThread "Reference Handler" daemon [_thread_blocked, id=4696, stack(0x01640000,0x01690000)] =0x016b9000 JavaThread "main" [_thread_in_vm, id=900, stack(0x00160000,0x001b0000)] Other Threads: 0x01751c00 VMThread [stack: 0x015f0000,0x01640000] [id=4052] 0x0179c800 WatcherThread [stack: 0x0b950000,0x0b9a0000] [id=3340] VM state:not at safepoint (normal execution) VM Mutex/Monitor currently owned by a thread: None Heap def new generation total 960K, used 816K [0x037c0000, 0x038c0000, 0x03ca0000) eden space 896K, 91% used [0x037c0000, 0x0388c2c0, 0x038a0000) from space 64K, 0% used [0x038a0000, 0x038a0000, 0x038b0000) to space 64K, 0% used [0x038b0000, 0x038b0000, 0x038c0000) tenured generation total 4096K, used 0K [0x03ca0000, 0x040a0000, 0x077c0000) the space 4096K, 0% used [0x03ca0000, 0x03ca0000, 0x03ca0200, 0x040a0000) compacting perm gen total 12288K, used 2143K [0x077c0000, 0x083c0000, 0x0b7c0000) the space 12288K, 17% used [0x077c0000, 0x079d7e38, 0x079d8000, 0x083c0000) No shared spaces configured. Dynamic libraries: 0x00400000 - 0x00424000 C:\Program Files\Java\jre6\bin\javaw.exe 0x77550000 - 0x7768e000 C:\Windows\SYSTEM32\ntdll.dll 0x75a80000 - 0x75b54000 C:\Windows\system32\kernel32.dll 0x758d0000 - 0x7591b000 C:\Windows\system32\KERNELBASE.dll 0x759e0000 - 0x75a80000 C:\Windows\system32\ADVAPI32.dll 0x76070000 - 0x7611c000 C:\Windows\system32\msvcrt.dll 0x77250000 - 0x77269000 C:\Windows\SYSTEM32\sechost.dll 0x771a0000 - 0x77241000 C:\Windows\system32\RPCRT4.dll 0x76eb0000 - 0x76f79000 C:\Windows\system32\USER32.dll 0x76e60000 - 0x76eae000 C:\Windows\system32\GDI32.dll 0x77770000 - 0x7777a000 C:\Windows\system32\LPK.dll 0x75fd0000 - 0x7606e000 C:\Windows\system32\USP10.dll 0x770b0000 - 0x770cf000 C:\Windows\system32\IMM32.DLL 0x770d0000 - 0x7719c000 C:\Windows\system32\MSCTF.dll 0x7c340000 - 0x7c396000 C:\Program Files\Java\jre6\bin\msvcr71.dll 0x6d800000 - 0x6da8b000 C:\Program Files\Java\jre6\bin\client\jvm.dll 0x73a00000 - 0x73a32000 C:\Windows\system32\WINMM.dll 0x75610000 - 0x7565b000 C:\Windows\system32\apphelp.dll 0x6d7b0000 - 0x6d7bc000 C:\Program Files\Java\jre6\bin\verify.dll 0x6d330000 - 0x6d34f000 C:\Program Files\Java\jre6\bin\java.dll 0x6d290000 - 0x6d298000 C:\Program Files\Java\jre6\bin\hpi.dll 0x776e0000 - 0x776e5000 C:\Windows\system32\PSAPI.DLL 0x6d7f0000 - 0x6d7ff000 C:\Program Files\Java\jre6\bin\zip.dll 0x10000000 - 0x1004c000 C:\Users\theo\Desktop\workspace\JavaFX1\lib\natives\windows\lwjgl.dll 0x5d170000 - 0x5d238000 C:\Windows\system32\OPENGL32.dll 0x6e7b0000 - 0x6e7d2000 C:\Windows\system32\GLU32.dll 0x70620000 - 0x70707000 C:\Windows\system32\DDRAW.dll 0x70610000 - 0x70616000 C:\Windows\system32\DCIMAN32.dll 0x75b60000 - 0x75cfd000 C:\Windows\system32\SETUPAPI.dll 0x759b0000 - 0x759d7000 C:\Windows\system32\CFGMGR32.dll 0x76d70000 - 0x76dff000 C:\Windows\system32\OLEAUT32.dll 0x75db0000 - 0x75f0c000 C:\Windows\system32\ole32.dll 0x758b0000 - 0x758c2000 C:\Windows\system32\DEVOBJ.dll 0x74060000 - 0x74073000 C:\Windows\system32\dwmapi.dll 0x74b60000 - 0x74b69000 C:\Windows\system32\VERSION.dll 0x745f0000 - 0x7478e000 C:\Windows\WinSxS\x86_microsoft.windows.common-controls_6595b64144ccf1df_6.0.7600.16661_none_420fe3fa2b8113bd\COMCTL32.dll 0x75d50000 - 0x75da7000 C:\Windows\system32\SHLWAPI.dll 0x74370000 - 0x743b0000 C:\Windows\system32\uxtheme.dll 0x22200000 - 0x22206000 C:\Program Files\ESET\ESET Smart Security\eplgHooks.dll VM Arguments: jvm_args: -Djava.library.path=C:\Users\theo\Desktop\workspace\JavaFX1\lib\natives\windows -Dfile.encoding=Cp1253 java_command: zarkopafilis.koding.io.javafx.Main Launcher Type: SUN_STANDARD Environment Variables: PATH=C:/Program Files/Java/jre6/bin/client;C:/Program Files/Java/jre6/bin;C:/Program Files/Java/jre6/lib/i386;C:\Perl\site\bin;C:\Perl\bin;C:\Ruby200\bin;C:\Program Files\Common Files\Microsoft Shared\Windows Live;C:\Windows\system32;C:\Windows;C:\Windows\System32\Wbem;C:\Windows\System32\WindowsPowerShell\v1.0\;C:\Program Files\Windows Live\Shared;C:\Users\theo\Desktop\eclipse; USERNAME=theo OS=Windows_NT PROCESSOR_IDENTIFIER=x86 Family 6 Model 37 Stepping 5, GenuineIntel --------------- S Y S T E M --------------- OS: Windows 7 Build 7600 CPU:total 4 (8 cores per cpu, 2 threads per core) family 6 model 37 stepping 5, cmov, cx8, fxsr, mmx, sse, sse2, sse3, ssse3, sse4.1, sse4.2, ht Memory: 4k page, physical 2097151k(1257972k free), swap 4194303k(4194303k free) vm_info: Java HotSpot(TM) Client VM (14.2-b01) for windows-x86 JRE (1.6.0_16-b01), built on Jul 31 2009 11:26:58 by "java_re" with MS VC++ 7.1 time: Wed Oct 23 22:00:12 2013 elapsed time: 0 seconds Code: Display.setDisplayMode(new DisplayMode(800,600)); Display.create();//Error here I am using JDK 6

  Read the article

• Geek Fun: Virtualized Old School Windows – Windows 95

  - by Matthew Guay

  Last week we enjoyed looking at Windows 3.1 running in VMware Player on Windows 7.  Today, let’s upgrade our 3.1 to 95, and get a look at how most of us remember Windows from the 90’s. In this demo, we’re running the first release of Windows 95 (version 4.00.950) in VMware Player 3.0 running on Windows 7 x64.  For fun, we ran the 95 upgrade on the 3.1 virtual machine we built last week. Windows 95 So let’s get started.  Here’s the first setup screen.  For the record, Windows 95 installed in about 15 minutes or less in VMware in our test. Strangely, Windows 95 offered several installation choices.  They actually let you choose what extra parts of Windows to install if you wished.  Oh, and who wants to run Windows 95 on your “Portable Computer”?  Most smartphones today are more powerful than the “portable computers” of 95. Your productivity may vastly increase if you run Windows 95.  Anyone want to switch? No, I don’t want to restart … I want to use my computer! Welcome to Windows 95!  Hey, did you know you can launch programs from the Start button? Our quick spin around Windows 95 reminded us why Windows got such a bad reputation in the ‘90’s for being unstable.  We didn’t even get our test copy fully booted after installation before we saw our first error screen.  Windows in space … was that the most popular screensaver in Windows 95, or was it just me? Hello Windows 3.1!  The UI was still outdated in some spots.   Ah, yes, Media Player before it got 101 features to compete with iTunes. But, you couldn’t even play CDs in Media Player.  Actually, CD player was one program I used almost daily in Windows 95 back in the day. Want some new programs?  This help file about new programs designed for Windows 95 lists a lot of outdated names in tech.    And, you really may want some programs.  The first edition of Windows 95 didn’t even ship with Internet Explorer.   We’ve still got Minesweeper, though! My Computer had really limited functionality, and by default opened everything in a new window.  Double click on C:, and it opens in a new window.  Ugh. But Explorer is a bit more like more modern versions. Hey, look, Start menu search!  If only it found the files you were looking for… Now I’m feeling old … this shutdown screen brought back so many memories … of shutdowns that wouldn’t shut down! But, you still have to turn off your computer.  I wonder how many old monitors had these words burned into them? So there’s yet another trip down Windows memory lane.  Most of us can remember using Windows 95, so let us know your favorite (or worst) memory of it!  At least we can all be thankful for our modern computers and operating systems today, right?  Similar Articles Productive Geek Tips Geek Fun: Remember the Old-School SkiFree Game?Geek Fun: Virtualized old school Windows 3.11Stupid Geek Tricks: Tile or Cascade Multiple Windows in Windows 7Stupid Geek Tricks: Select Multiple Windows on the TaskbarHow to Delete a System File in Windows 7 or Vista TouchFreeze Alternative in AutoHotkey The Icy Undertow Desktop Windows Home Server – Backup to LAN The Clear & Clean Desktop Use This Bookmarklet to Easily Get Albums Use AutoHotkey to Assign a Hotkey to a Specific Window Latest Software Reviews Tinyhacker Random Tips Revo Uninstaller Pro Registry Mechanic 9 for Windows PC Tools Internet Security Suite 2010 PCmover Professional Enable Check Box Selection in Windows 7 OnlineOCR – Free OCR Service Betting on the Blind Side, a Vanity Fair article 30 Minimal Logo Designs that Say More with Less LEGO Digital Designer – Free Create a Personal Website Quickly using Flavors.me

  Read the article

• Geek Fun: Virtualized Old School Windows – Windows 95

  - by Matthew Guay

  Last week we enjoyed looking at Windows 3.1 running in VMware Player on Windows 7.  Today, let’s upgrade our 3.1 to 95, and get a look at how most of us remember Windows from the 90’s. In this demo, we’re running the first release of Windows 95 (version 4.00.950) in VMware Player 3.0 running on Windows 7 x64.  For fun, we ran the 95 upgrade on the 3.1 virtual machine we built last week. Windows 95 So let’s get started.  Here’s the first setup screen.  For the record, Windows 95 installed in about 15 minutes or less in VMware in our test. Strangely, Windows 95 offered several installation choices.  They actually let you choose what extra parts of Windows to install if you wished.  Oh, and who wants to run Windows 95 on your “Portable Computer”?  Most smartphones today are more powerful than the “portable computers” of 95. Your productivity may vastly increase if you run Windows 95.  Anyone want to switch? No, I don’t want to restart … I want to use my computer! Welcome to Windows 95!  Hey, did you know you can launch programs from the Start button? Our quick spin around Windows 95 reminded us why Windows got such a bad reputation in the ‘90’s for being unstable.  We didn’t even get our test copy fully booted after installation before we saw our first error screen.  Windows in space … was that the most popular screensaver in Windows 95, or was it just me? Hello Windows 3.1!  The UI was still outdated in some spots.   Ah, yes, Media Player before it got 101 features to compete with iTunes. But, you couldn’t even play CDs in Media Player.  Actually, CD player was one program I used almost daily in Windows 95 back in the day. Want some new programs?  This help file about new programs designed for Windows 95 lists a lot of outdated names in tech.    And, you really may want some programs.  The first edition of Windows 95 didn’t even ship with Internet Explorer.   We’ve still got Minesweeper, though! My Computer had really limited functionality, and by default opened everything in a new window.  Double click on C:, and it opens in a new window.  Ugh. But Explorer is a bit more like more modern versions. Hey, look, Start menu search!  If only it found the files you were looking for… Now I’m feeling old … this shutdown screen brought back so many memories … of shutdowns that wouldn’t shut down! But, you still have to turn off your computer.  I wonder how many old monitors had these words burned into them? So there’s yet another trip down Windows memory lane.  Most of us can remember using Windows 95, so let us know your favorite (or worst) memory of it!  At least we can all be thankful for our modern computers and operating systems today, right?  Similar Articles Productive Geek Tips Geek Fun: Remember the Old-School SkiFree Game?Geek Fun: Virtualized old school Windows 3.11Stupid Geek Tricks: Tile or Cascade Multiple Windows in Windows 7Stupid Geek Tricks: Select Multiple Windows on the TaskbarHow to Delete a System File in Windows 7 or Vista TouchFreeze Alternative in AutoHotkey The Icy Undertow Desktop Windows Home Server – Backup to LAN The Clear & Clean Desktop Use This Bookmarklet to Easily Get Albums Use AutoHotkey to Assign a Hotkey to a Specific Window Latest Software Reviews Tinyhacker Random Tips Revo Uninstaller Pro Registry Mechanic 9 for Windows PC Tools Internet Security Suite 2010 PCmover Professional Enable Check Box Selection in Windows 7 OnlineOCR – Free OCR Service Betting on the Blind Side, a Vanity Fair article 30 Minimal Logo Designs that Say More with Less LEGO Digital Designer – Free Create a Personal Website Quickly using Flavors.me

  Read the article

• Unity Locks Up in Live CD

  - by user212883

  I'm trying to run from the live USB to install Ubuntu 13.10 on my Windows Machine (as I've grown a touch sick of Windows). However, whenever I boot into the LiveUSB session after a few moments the Unity desktop locks up (except the mouse pointer, which I can move). Is this something to do with the fact I've got an NVidia 580 GTX? I've heard of issues with Ubuntu and this card. I've also got an SSD, but given that it's booting from USB I shouldn't think that's an issue. System Specs: Processor: Intel Core i7-2600K CPU @ 3.40 GHZ Motherboard: Asus Maximus IV Gene-Z Z68 Socket 1155 RAM: 8GB DDR3 GPU: ASUS NVidia 580 GTX

  Read the article

• Understanding and Controlling Parallel Query Processing in SQL Server

  Data warehousing and general reporting applications tend to be CPU intensive because they need to read and process a large number of rows. To facilitate quick data processing for queries that touch a large amount of data, Microsoft SQL Server exploits the power of multiple logical processors to provide parallel query processing operations such as parallel scans. Through extensive testing, we have learned that, for most large queries that are executed in a parallel fashion, SQL Server can deliver linear or nearly linear response time speedup as the number of logical processors increases. However, some queries in high parallelism scenarios perform suboptimally. There are also some parallelism issues that can occur in a multi-user parallel query workload. This white paper describes parallel performance problems you might encounter when you run such queries and workloads, and it explains why these issues occur. In addition, it presents how data warehouse developers can detect these issues, and how they can work around them or mitigate them.

  Read the article

• Update Manager unable to get updates

  - by dPEN

  In last few days my Ubuntu 11.10 update manager is unable to get new updates. When I checked update log I saw that for couple of updates it says "Network isn't available". For other updates it downloaded logs and and internet connection also works fine. Unable to attached screenshot due to SPAM prevention policy. But for below Release gpgv:/var/lib/apt/lists/partial/extras.ubuntu.com_ubuntu_dists_oneiric_Release.gpg it says "Network isn't available" For all other Releases it is downloading fine. And due to this I dont see any update available in last 10 days. LOG OF sudo apt-get update: dipen@EIDLCPU1018:~$ sudo apt-get update [sudo] password for dipen: Ign http:/extras.ubuntu.com oneiric InRelease Ign http:/archive.canonical.com oneiric InRelease Ign http:/archive.canonical.com lucid InRelease Get:1 http:/extras.ubuntu.com oneiric Release.gpg [72 B] Get:2 http:/archive.canonical.com oneiric Release.gpg [198 B] Hit http:/extras.ubuntu.com oneiric Release Get:3 http:/archive.canonical.com lucid Release.gpg [198 B] Err http:/extras.ubuntu.com oneiric Release Hit http:/archive.canonical.com oneiric Release Ign http:/archive.canonical.com oneiric Release Hit http:/archive.canonical.com lucid Release Ign http:/archive.canonical.com lucid Release Ign http:/archive.canonical.com oneiric/partner i386 Packages/DiffIndex Ign http:/archive.canonical.com oneiric/partner TranslationIndex Ign http:/archive.canonical.com lucid/partner i386 Packages/DiffIndex Ign http:/archive.canonical.com lucid/partner TranslationIndex Hit http:/archive.canonical.com oneiric/partner i386 Packages Hit http:/archive.canonical.com lucid/partner i386 Packages Ign http:/dl.google.com stable InRelease Ign http:/archive.canonical.com oneiric/partner Translation-en_IN Ign http:/archive.canonical.com oneiric/partner Translation-en Ign http:/archive.canonical.com lucid/partner Translation-en_IN Ign http:/archive.canonical.com lucid/partner Translation-en Ign http:/in.archive.ubuntu.com oneiric InRelease Ign http:/in.archive.ubuntu.com oneiric-updates InRelease Ign http:/in.archive.ubuntu.com oneiric-security InRelease Get:4 http//dl.google.com stable Release.gpg [198 B] Get:5 http//in.archive.ubuntu.com oneiric Release.gpg [198 B] Get:6 http//in.archive.ubuntu.com oneiric-updates Release.gpg [198 B] Get:7 http//in.archive.ubuntu.com oneiric-security Release.gpg [198 B] Hit http:/in.archive.ubuntu.com oneiric Release Ign http:/in.archive.ubuntu.com oneiric Release Hit http:/in.archive.ubuntu.com oneiric-updates Release Err http:/in.archive.ubuntu.com oneiric-updates Release Hit http:/in.archive.ubuntu.com oneiric-security Release Ign http:/in.archive.ubuntu.com oneiric-security Release Ign http:/in.archive.ubuntu.com oneiric/main i386 Packages/DiffIndex Ign http:/in.archive.ubuntu.com oneiric/restricted i386 Packages/DiffIndex Ign http:/in.archive.ubuntu.com oneiric/universe i386 Packages/DiffIndex Ign http:/in.archive.ubuntu.com oneiric/multiverse i386 Packages/DiffIndex Hit http:/in.archive.ubuntu.com oneiric/main TranslationIndex Hit http:/in.archive.ubuntu.com oneiric/multiverse TranslationIndex Hit http:/in.archive.ubuntu.com oneiric/restricted TranslationIndex Hit http:/in.archive.ubuntu.com oneiric/universe TranslationIndex Ign http:/in.archive.ubuntu.com oneiric-security/main i386 Packages/DiffIndex Ign http:/in.archive.ubuntu.com oneiric-security/restricted i386 Packages/DiffIndex Ign http:/in.archive.ubuntu.com oneiric-security/universe i386 Packages/DiffIndex Ign http:/in.archive.ubuntu.com oneiric-security/multiverse i386 Packages/DiffIndex Hit http:/in.archive.ubuntu.com oneiric-security/main TranslationIndex Hit http:/in.archive.ubuntu.com oneiric-security/multiverse TranslationIndex Hit http:/in.archive.ubuntu.com oneiric-security/restricted TranslationIndex Hit http:/in.archive.ubuntu.com oneiric-security/universe TranslationIndex Hit http:/in.archive.ubuntu.com oneiric/main i386 Packages Hit http:/in.archive.ubuntu.com oneiric/restricted i386 Packages Hit http:/in.archive.ubuntu.com oneiric/universe i386 Packages Hit http:/in.archive.ubuntu.com oneiric/multiverse i386 Packages Hit http:/in.archive.ubuntu.com oneiric/main Translation-en Hit http:/in.archive.ubuntu.com oneiric/multiverse Translation-en Hit http:/in.archive.ubuntu.com oneiric/restricted Translation-en Hit http:/in.archive.ubuntu.com oneiric/universe Translation-en Hit http:/in.archive.ubuntu.com oneiric-security/main i386 Packages Hit http:/in.archive.ubuntu.com oneiric-security/restricted i386 Packages Hit http:/in.archive.ubuntu.com oneiric-security/universe i386 Packages Hit http:/in.archive.ubuntu.com oneiric-security/multiverse i386 Packages Hit http:/in.archive.ubuntu.com oneiric-security/main Translation-en Hit http:/in.archive.ubuntu.com oneiric-security/multiverse Translation-en Get:8 http//dl.google.com stable Release [1,347 B] Hit http:/in.archive.ubuntu.com oneiric-security/restricted Translation-en Hit http:/in.archive.ubuntu.com oneiric-security/universe Translation-en Get:9 http//dl.google.com stable/main i386 Packages [1,214 B] Ign http:/dl.google.com stable/main TranslationIndex Ign http:/dl.google.com stable/main Translation-en_IN Ign http:/dl.google.com stable/main Translation-en Fetched 3,821 B in 41s (91 B/s) Reading package lists... Done W: A error occurred during the signature verification. The repository is not updated and the previous index files will be used. GPG error: http:/extras.ubuntu.com oneiric Release: The following signatures were invalid: BADSIG 16126D3A3E5C1192 Ubuntu Extras Archive Automatic Signing Key <[email protected]> W: GPG error: http:/archive.canonical.com oneiric Release: The following signatures were invalid: BADSIG 40976EAF437D05B5 Ubuntu Archive Automatic Signing Key <[email protected]> W: GPG error: http:/archive.canonical.com lucid Release: The following signatures were invalid: BADSIG 40976EAF437D05B5 Ubuntu Archive Automatic Signing Key <[email protected]> W: GPG error: http:/in.archive.ubuntu.com oneiric Release: The following signatures were invalid: BADSIG 40976EAF437D05B5 Ubuntu Archive Automatic Signing Key <[email protected]> W: A error occurred during the signature verification. The repository is not updated and the previous index files will be used. GPG error: http:/in.archive.ubuntu.com oneiric-updates Release: The following signatures were invalid: BADSIG 40976EAF437D05B5 Ubuntu Archive Automatic Signing Key <[email protected]> W: GPG error: http:/in.archive.ubuntu.com oneiric-security Release: The following signatures were invalid: BADSIG 40976EAF437D05B5 Ubuntu Archive Automatic Signing Key <[email protected]> W: Failed to fetch http:/extras.ubuntu.com/ubuntu/dists/oneiric/Release W: Failed to fetch http:/in.archive.ubuntu.com/ubuntu/dists/oneiric-updates/Release W: Some index files failed to download. They have been ignored, or old ones used instead. dipen@EIDLCPU1018:~$ LOG of sudo apt-get upgrade: dipen@EIDLCPU1018:~$ sudo apt-get upgrade Reading package lists... Done Building dependency tree Reading state information... Done The following packages have been kept back: ghc6-doc haskell-zlib-doc libghc6-zlib-doc 0 upgraded, 0 newly installed, 0 to remove and 3 not upgraded. dipen@EIDLCPU1018:~$ /etc/apt/sources.list: deb http:/in.archive.ubuntu.com/ubuntu/ oneiric main restricted deb http:/in.archive.ubuntu.com/ubuntu/ oneiric-updates main restricted deb http:/in.archive.ubuntu.com/ubuntu/ oneiric universe deb http:/in.archive.ubuntu.com/ubuntu/ oneiric-updates universe deb http:/in.archive.ubuntu.com/ubuntu/ oneiric multiverse deb http:/in.archive.ubuntu.com/ubuntu/ oneiric-updates multiverse deb http:/archive.canonical.com/ubuntu oneiric partner deb http:/in.archive.ubuntu.com/ubuntu/ oneiric-security main restricted deb http:/in.archive.ubuntu.com/ubuntu/ oneiric-security universe deb http:/in.archive.ubuntu.com/ubuntu/ oneiric-security multiverse deb http:/extras.ubuntu.com/ubuntu oneiric main #Third party developers repository deb http:/archive.canonical.com/ lucid partner

  Read the article

• Database Vault 11gR2 11.2.0.1 Certified with Oracle E-Business Suite

  - by Steven Chan

  Oracle Database Vault allows security administrators to protect a database from privileged account access to application data.  Database objects can be placed in protected realms, which can be accessed only if a specific set of conditions are met.  Oracle Database Vault 11gR2 11.2.0.1 is now certified with Oracle E-Business Suite Release 11i and 12.You can now enable Database Vault 11gR2 on your existing E-Business Suite 11.2.0.1 Database instance.  If you already have DB Vault 10gR2 or 11gR1 enabled in your E-Business Suite environment, you can now upgrade to the 11gR2 Database.  We also support EBS patching with Database Vault 11.2.0.1 enabled. Our DB Vault realm creation and grants-related scripts have been updated to reduce patching downtimes.

  Read the article

• Difference between 12.04 and 12.04.1

  - by Jeff

  I recently did a fresh install of Ubuntu 12.04 two days ago. Or at least I thought it was 12.04, but actually 12.04.1. Now I'm having errors popping up from the grub loader. Error: no video mode activated which was apparently resolved in this bug# 699802. However these workarounds are for 11.xx and not working for me. I never had these errors before with 12.04 and now I'm getting them. What's the difference between 12.04 and 12.04.1? Off the bat I notice that the kernels are different 12.04 uses 3.2.0-26-generic-pae 12.04.1 uses 3.2.0-29-generic after an immediate sudo apt-get update upgrade 12.04.1 uses 3.2.0-30-generic I have two other computers running 12.04 (not 12.04.1) and they're working fine. The computer that I'm currently was working fine (with 12.04) previously too. Should I roll back my kernel to 3.2.0-26?

  Read the article

• Use a Windows 8-Like Task Manager in Windows 7, Vista, and XP

  - by Lori Kaufman

  One of the new features in Windows 8 is the improved Task Manager, which provides access to more information and settings. If you don’t want to upgrade, there is a way you can use a simple Windows 8-like Task Manager in Windows 7, Vista, or XP. The Windows 8 Metro Task Manager does not need to be installed. Simply download the .zip file (see the download link at the end of this article), extract the files, and double-click the Windows 8 Task Manager.exe file. A window displays a list of tasks currently running with the status of each task listed. To end a task, select the task in the list and click End Task. Why Does 64-Bit Windows Need a Separate “Program Files (x86)” Folder? Why Your Android Phone Isn’t Getting Operating System Updates and What You Can Do About It How To Delete, Move, or Rename Locked Files in Windows

  Read the article

• New Whitepaper: Upgrading EBS 11i Forms + OA Framework Personalizations to EBS 12

  - by Sara Woodhull

  Personalizations are -- and have always been -- one of the safest and most upgradable ways to "customize" your Oracle E-Business Suite screens, both for Oracle Forms-based screens and for Oracle Application Framework-based pages. However, the upgrade from Release 11i to Release 12.1 spans many years of EBS evolution, during which time Oracle has actively been building many new features and modules. A lot has changed in Oracle E-Business Suite that may affect upgrading your personalizations from 11i to 12.1. We have published a new note on My Oracle Support that discusses ways to evaluate your existing personalizations:Upgrading Form Personalizations and OA Framework Personalizations from Oracle E-Business Suite Release 11i to 12.1 (Note 1292611.1)Two distinct types of personalizations There are two distinct types of personalizations: Form Personalization OA Framework Personalization. Both types of personalization are completely metadata-based. The personalizations are stored as data in database tables. However, because the underlying technologies (Oracle Forms and OA Framework) are very different, Forms personalizations and OA Framework personalizations are not equivalent and cannot be converted or migrated from one to the other.

  Read the article

• How to install Juniper VPN on Ubuntu 14.04 LTS?

  - by Max Ricardo Mercurio Ribeiro

  Could you please help me ? On my old Ubuntu 13.10 I was able to run Juniper VPN (on Firefox only) using a workaround which requires you to install the missing 32libs and IcedTea (32bits). However, I recently upgraded from Ubuntu 13.10 to 14.04 (both 64 bits) and my Juniper VPN does not work anymore because it fails during startup showing the following message: "Please ensure that necessary 32 bit libraries are installed. For more details, refer KB article KB25230" "Setup failed. Please install 32 bit Java and update alternatives links using update-alternatives command. For more details, refer KB article KB25230" For some odd reason, it seems the 14.04 upgrade do not work anymore with the openjdk-7:386 and consequently the Juniper VPN as well. Any ideas ? Thanks

  Read the article

• Identifying Data Model Changes Between EBS 12.1.3 and Prior EBS Releases

  - by Steven Chan

  The EBS 12.1.3 Release Content Document (RCD, Note 561580.1) summarizes the latest functional and technology stack-related updates in a specific release.  The E-Business Suite Electronic Technical Reference Manual (eTRM) summarizes the database objects in a specific EBS release.  Those are useful references, but sometimes you need to find out which database objects have changed between one EBS release and another.  This kind of information about the differences or deltas between two releases is useful if you have customized or extended your EBS instance and plan to upgrade to EBS 12.1.3. Where can you find that information?Answering that question has just gotten a lot easier.  You can now use a new EBS Data Model Comparison Report tool:EBS Data Model Comparison Report Overview (Note 1290886.1)This new tool lists the database object definition changes between the following source and target EBS releases:EBS 11.5.10.2 and EBS 12.1.3EBS 12.0.4 and EBS 12.1.3EBS 12.1.1 and EBS 12.1.3EBS 12.1.2 and EBS 12.1.3For example, here's part of the report comparing Bill of Materials changes between 11.5.10.2 and 12.1.3:

  Read the article

• SQL SERVER – Database Dynamic Caching by Automatic SQL Server Performance Acceleration

  - by pinaldave

  My second look at SafePeak’s new version (2.1) revealed to me few additional interesting features. For those of you who hadn’t read my previous reviews SafePeak and not familiar with it, here is a quick brief: SafePeak is in business of accelerating performance of SQL Server applications, as well as their scalability, without making code changes to the applications or to the databases. SafePeak performs database dynamic caching, by caching in memory result sets of queries and stored procedures while keeping all those cache correct and up to date. Cached queries are retrieved from the SafePeak RAM in microsecond speed and not send to the SQL Server. The application gets much faster results (100-500 micro seconds), the load on the SQL Server is reduced (less CPU and IO) and the application or the infrastructure gets better scalability. SafePeak solution is hosted either within your cloud servers, hosted servers or your enterprise servers, as part of the application architecture. Connection of the application is done via change of connection strings or adding reroute line in the c:\windows\system32\drivers\etc\hosts file on all application servers. For those who would like to learn more on SafePeak architecture and how it works, I suggest to read this vendor’s webpage: SafePeak Architecture. More interesting new features in SafePeak 2.1 In my previous review of SafePeak new I covered the first 4 things I noticed in the new SafePeak (check out my article “SQLAuthority News – SafePeak Releases a Major Update: SafePeak version 2.1 for SQL Server Performance Acceleration”): Cache setup and fine-tuning – a critical part for getting good caching results Database templates Choosing which database to cache Monitoring and analysis options by SafePeak Since then I had a chance to play with SafePeak some more and here is what I found. 5. Analysis of SQL Performance (present and history): In SafePeak v.2.1 the tools for understanding of performance became more comprehensive. Every 15 minutes SafePeak creates and updates various performance statistics. Each query (or a procedure execute) that arrives to SafePeak gets a SQL pattern, and after it is used again there are statistics for such pattern. An important part of this product is that it understands the dependencies of every pattern (list of tables, views, user defined functions and procs). From this understanding SafePeak creates important analysis information on performance of every object: response time from the database, response time from SafePeak cache, average response time, percent of traffic and break down of behavior. One of the interesting things this behavior column shows is how often the object is actually pdated. The break down analysis allows knowing the above information for: queries and procedures, tables, views, databases and even instances level. The data is show now on all arriving queries, both read queries (that can be cached), but also any types of updates like DMLs, DDLs, DCLs, and even session settings queries. The stats are being updated every 15 minutes and SafePeak dashboard allows going back in time and investigating what happened within any time frame. 6. Logon trigger, for making sure nothing corrupts SafePeak cache data If you have an application with many parts, many servers many possible locations that can actually update the database, or the SQL Server is accessible to many DBAs or software engineers, each can access some database directly and do some changes without going thru SafePeak – this can create a potential corruption of the data stored in SafePeak cache. To make sure SafePeak cache is correct it needs to get all updates to arrive to SafePeak, and if a DBA will access the database directly and do some changes, for example, then SafePeak will simply not know about it and will not clean SafePeak cache. In the new version, SafePeak brought a new feature called “Logon Trigger” to solve the above challenge. By special click of a button SafePeak can deploy a special server logon trigger (with a CLR object) on your SQL Server that actually monitors all connections and informs SafePeak on any connection that is coming not from SafePeak. In SafePeak dashboard there is an interface that allows to control which logins can be ignored based on login names and IPs, while the rest will invoke cache cleanup of SafePeak and actually locks SafePeak cache until this connection will not be closed. Important to note, that this does not interrupt any logins, only informs SafePeak on such connection. On the Dashboard screen in SafePeak you will be able to see those connections and then decide what to do with them. Configuration of this feature in SafePeak dashboard can be done here: Settings -> SQL instances management -> click on instance -> Logon Trigger tab. Other features: 7. User management ability to grant permissions to someone without changing its configuration and only use SafePeak as performance analysis tool. 8. Better reports for analysis of performance using 15 minute resolution charts. 9. Caching of client cursors 10. Support for IPv6 Summary SafePeak is a great SQL Server performance acceleration solution for users who want immediate results for sites with performance, scalability and peak spikes challenges. Especially if your apps are packaged or 3rd party, since no code changes are done. SafePeak can significantly increase response times, by reducing network roundtrip to the database, decreasing CPU resource usage, eliminating I/O and storage access. SafePeak team provides a free fully functional trial www.safepeak.com/download and actually provides a one-on-one assistance during such trial. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: About Me, Pinal Dave, PostADay, SQL, SQL Authority, SQL Performance, SQL Query, SQL Server, SQL Tips and Tricks, SQL Utility, T SQL, Technology

  Read the article

• What a Performance! MySQL 5.5 and InnoDB 1.1 running on Oracle Linux

  - by zeynep.koch(at)oracle.com

  The MySQL performance team in Oracle has recently completed a series of benchmarks comparing Read / Write and Read-Only performance of MySQL 5.5 with the InnoDB and MyISAM storage engines. Compared to MyISAM, InnoDB delivered 35x higher throughput on the Read / Write test and 5x higher throughput on the Read-Only test, with 90% scalability across 36 CPU cores. A full analysis of results and MySQL configuration parameters are documented in a new whitepaperIn addition to the benchmark, the new whitepaper, also includes:- A discussion of the use-cases for each storage engine- Best practices for users considering the migration of existing applications from MyISAM to InnoDB- A summary of the performance and scalability enhancements introduced with MySQL 5.5 and InnoDB 1.1.The benchmark itself was based on Sysbench, running on AMD Opteron "Magny-Cours" processors, and Oracle Linux with the Unbreakable Enterprise Kernel You can learn more about MySQL 5.5 and InnoDB 1.1 from here and download it from here to test whether you witness performance gains in your real-world applications.  By Mat Keep

  Read the article

• Improving Partitioned Table Join Performance

  - by Paul White

  The query optimizer does not always choose an optimal strategy when joining partitioned tables. This post looks at an example, showing how a manual rewrite of the query can almost double performance, while reducing the memory grant to almost nothing. Test Data The two tables in this example use a common partitioning partition scheme. The partition function uses 41 equal-size partitions: CREATE PARTITION FUNCTION PFT (integer) AS RANGE RIGHT FOR VALUES ( 125000, 250000, 375000, 500000, 625000, 750000, 875000, 1000000, 1125000, 1250000, 1375000, 1500000, 1625000, 1750000, 1875000, 2000000, 2125000, 2250000, 2375000, 2500000, 2625000, 2750000, 2875000, 3000000, 3125000, 3250000, 3375000, 3500000, 3625000, 3750000, 3875000, 4000000, 4125000, 4250000, 4375000, 4500000, 4625000, 4750000, 4875000, 5000000 ); GO CREATE PARTITION SCHEME PST AS PARTITION PFT ALL TO ([PRIMARY]); There two tables are: CREATE TABLE dbo.T1 ( TID integer NOT NULL IDENTITY(0,1), Column1 integer NOT NULL, Padding binary(100) NOT NULL DEFAULT 0x,   CONSTRAINT PK_T1 PRIMARY KEY CLUSTERED (TID) ON PST (TID) );   CREATE TABLE dbo.T2 ( TID integer NOT NULL, Column1 integer NOT NULL, Padding binary(100) NOT NULL DEFAULT 0x,   CONSTRAINT PK_T2 PRIMARY KEY CLUSTERED (TID, Column1) ON PST (TID) ); The next script loads 5 million rows into T1 with a pseudo-random value between 1 and 5 for Column1. The table is partitioned on the IDENTITY column TID: INSERT dbo.T1 WITH (TABLOCKX) (Column1) SELECT (ABS(CHECKSUM(NEWID())) % 5) + 1 FROM dbo.Numbers AS N WHERE n BETWEEN 1 AND 5000000; In case you don’t already have an auxiliary table of numbers lying around, here’s a script to create one with 10 million rows: CREATE TABLE dbo.Numbers (n bigint PRIMARY KEY);   WITH L0 AS(SELECT 1 AS c UNION ALL SELECT 1), L1 AS(SELECT 1 AS c FROM L0 AS A CROSS JOIN L0 AS B), L2 AS(SELECT 1 AS c FROM L1 AS A CROSS JOIN L1 AS B), L3 AS(SELECT 1 AS c FROM L2 AS A CROSS JOIN L2 AS B), L4 AS(SELECT 1 AS c FROM L3 AS A CROSS JOIN L3 AS B), L5 AS(SELECT 1 AS c FROM L4 AS A CROSS JOIN L4 AS B), Nums AS(SELECT ROW_NUMBER() OVER (ORDER BY (SELECT NULL)) AS n FROM L5) INSERT dbo.Numbers WITH (TABLOCKX) SELECT TOP (10000000) n FROM Nums ORDER BY n OPTION (MAXDOP 1); Table T1 contains data like this: Next we load data into table T2. The relationship between the two tables is that table 2 contains ‘n’ rows for each row in table 1, where ‘n’ is determined by the value in Column1 of table T1. There is nothing particularly special about the data or distribution, by the way. INSERT dbo.T2 WITH (TABLOCKX) (TID, Column1) SELECT T.TID, N.n FROM dbo.T1 AS T JOIN dbo.Numbers AS N ON N.n >= 1 AND N.n <= T.Column1; Table T2 ends up containing about 15 million rows: The primary key for table T2 is a combination of TID and Column1. The data is partitioned according to the value in column TID alone. Partition Distribution The following query shows the number of rows in each partition of table T1: SELECT PartitionID = CA1.P, NumRows = COUNT_BIG(*) FROM dbo.T1 AS T CROSS APPLY (VALUES ($PARTITION.PFT(TID))) AS CA1 (P) GROUP BY CA1.P ORDER BY CA1.P; There are 40 partitions containing 125,000 rows (40 * 125k = 5m rows). The rightmost partition remains empty. The next query shows the distribution for table 2: SELECT PartitionID = CA1.P, NumRows = COUNT_BIG(*) FROM dbo.T2 AS T CROSS APPLY (VALUES ($PARTITION.PFT(TID))) AS CA1 (P) GROUP BY CA1.P ORDER BY CA1.P; There are roughly 375,000 rows in each partition (the rightmost partition is also empty): Ok, that’s the test data done. Test Query and Execution Plan The task is to count the rows resulting from joining tables 1 and 2 on the TID column: SET STATISTICS IO ON; DECLARE @s datetime2 = SYSUTCDATETIME();   SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID;   SELECT DATEDIFF(Millisecond, @s, SYSUTCDATETIME()); SET STATISTICS IO OFF; The optimizer chooses a plan using parallel hash join, and partial aggregation: The Plan Explorer plan tree view shows accurate cardinality estimates and an even distribution of rows across threads (click to enlarge the image): With a warm data cache, the STATISTICS IO output shows that no physical I/O was needed, and all 41 partitions were touched: Running the query without actual execution plan or STATISTICS IO information for maximum performance, the query returns in around 2600ms. Execution Plan Analysis The first step toward improving on the execution plan produced by the query optimizer is to understand how it works, at least in outline. The two parallel Clustered Index Scans use multiple threads to read rows from tables T1 and T2. Parallel scan uses a demand-based scheme where threads are given page(s) to scan from the table as needed. This arrangement has certain important advantages, but does result in an unpredictable distribution of rows amongst threads. The point is that multiple threads cooperate to scan the whole table, but it is impossible to predict which rows end up on which threads. For correct results from the parallel hash join, the execution plan has to ensure that rows from T1 and T2 that might join are processed on the same thread. For example, if a row from T1 with join key value ‘1234’ is placed in thread 5’s hash table, the execution plan must guarantee that any rows from T2 that also have join key value ‘1234’ probe thread 5’s hash table for matches. The way this guarantee is enforced in this parallel hash join plan is by repartitioning rows to threads after each parallel scan. The two repartitioning exchanges route rows to threads using a hash function over the hash join keys. The two repartitioning exchanges use the same hash function so rows from T1 and T2 with the same join key must end up on the same hash join thread. Expensive Exchanges This business of repartitioning rows between threads can be very expensive, especially if a large number of rows is involved. The execution plan selected by the optimizer moves 5 million rows through one repartitioning exchange and around 15 million across the other. As a first step toward removing these exchanges, consider the execution plan selected by the optimizer if we join just one partition from each table, disallowing parallelism: SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = 1 AND $PARTITION.PFT(T2.TID) = 1 OPTION (MAXDOP 1); The optimizer has chosen a (one-to-many) merge join instead of a hash join. The single-partition query completes in around 100ms. If everything scaled linearly, we would expect that extending this strategy to all 40 populated partitions would result in an execution time around 4000ms. Using parallelism could reduce that further, perhaps to be competitive with the parallel hash join chosen by the optimizer. This raises a question. If the most efficient way to join one partition from each of the tables is to use a merge join, why does the optimizer not choose a merge join for the full query? Forcing a Merge Join Let’s force the optimizer to use a merge join on the test query using a hint: SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID OPTION (MERGE JOIN); This is the execution plan selected by the optimizer: This plan results in the same number of logical reads reported previously, but instead of 2600ms the query takes 5000ms. The natural explanation for this drop in performance is that the merge join plan is only using a single thread, whereas the parallel hash join plan could use multiple threads. Parallel Merge Join We can get a parallel merge join plan using the same query hint as before, and adding trace flag 8649: SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID OPTION (MERGE JOIN, QUERYTRACEON 8649); The execution plan is: This looks promising. It uses a similar strategy to distribute work across threads as seen for the parallel hash join. In practice though, performance is disappointing. On a typical run, the parallel merge plan runs for around 8400ms; slower than the single-threaded merge join plan (5000ms) and much worse than the 2600ms for the parallel hash join. We seem to be going backwards! The logical reads for the parallel merge are still exactly the same as before, with no physical IOs. The cardinality estimates and thread distribution are also still very good (click to enlarge): A big clue to the reason for the poor performance is shown in the wait statistics (captured by Plan Explorer Pro): CXPACKET waits require careful interpretation, and are most often benign, but in this case excessive waiting occurs at the repartitioning exchanges. Unlike the parallel hash join, the repartitioning exchanges in this plan are order-preserving ‘merging’ exchanges (because merge join requires ordered inputs): Parallelism works best when threads can just grab any available unit of work and get on with processing it. Preserving order introduces inter-thread dependencies that can easily lead to significant waits occurring. In extreme cases, these dependencies can result in an intra-query deadlock, though the details of that will have to wait for another time to explore in detail. The potential for waits and deadlocks leads the query optimizer to cost parallel merge join relatively highly, especially as the degree of parallelism (DOP) increases. This high costing resulted in the optimizer choosing a serial merge join rather than parallel in this case. The test results certainly confirm its reasoning. Collocated Joins In SQL Server 2008 and later, the optimizer has another available strategy when joining tables that share a common partition scheme. This strategy is a collocated join, also known as as a per-partition join. It can be applied in both serial and parallel execution plans, though it is limited to 2-way joins in the current optimizer. Whether the optimizer chooses a collocated join or not depends on cost estimation. The primary benefits of a collocated join are that it eliminates an exchange and requires less memory, as we will see next. Costing and Plan Selection The query optimizer did consider a collocated join for our original query, but it was rejected on cost grounds. The parallel hash join with repartitioning exchanges appeared to be a cheaper option. There is no query hint to force a collocated join, so we have to mess with the costing framework to produce one for our test query. Pretending that IOs cost 50 times more than usual is enough to convince the optimizer to use collocated join with our test query: -- Pretend IOs are 50x cost temporarily DBCC SETIOWEIGHT(50);   -- Co-located hash join SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID OPTION (RECOMPILE);   -- Reset IO costing DBCC SETIOWEIGHT(1); Collocated Join Plan The estimated execution plan for the collocated join is: The Constant Scan contains one row for each partition of the shared partitioning scheme, from 1 to 41. The hash repartitioning exchanges seen previously are replaced by a single Distribute Streams exchange using Demand partitioning. Demand partitioning means that the next partition id is given to the next parallel thread that asks for one. My test machine has eight logical processors, and all are available for SQL Server to use. As a result, there are eight threads in the single parallel branch in this plan, each processing one partition from each table at a time. Once a thread finishes processing a partition, it grabs a new partition number from the Distribute Streams exchange…and so on until all partitions have been processed. It is important to understand that the parallel scans in this plan are different from the parallel hash join plan. Although the scans have the same parallelism icon, tables T1 and T2 are not being co-operatively scanned by multiple threads in the same way. Each thread reads a single partition of T1 and performs a hash match join with the same partition from table T2. The properties of the two Clustered Index Scans show a Seek Predicate (unusual for a scan!) limiting the rows to a single partition: The crucial point is that the join between T1 and T2 is on TID, and TID is the partitioning column for both tables. A thread that processes partition ‘n’ is guaranteed to see all rows that can possibly join on TID for that partition. In addition, no other thread will see rows from that partition, so this removes the need for repartitioning exchanges. CPU and Memory Efficiency Improvements The collocated join has removed two expensive repartitioning exchanges and added a single exchange processing 41 rows (one for each partition id). Remember, the parallel hash join plan exchanges had to process 5 million and 15 million rows. The amount of processor time spent on exchanges will be much lower in the collocated join plan. In addition, the collocated join plan has a maximum of 8 threads processing single partitions at any one time. The 41 partitions will all be processed eventually, but a new partition is not started until a thread asks for it. Threads can reuse hash table memory for the new partition. The parallel hash join plan also had 8 hash tables, but with all 5,000,000 build rows loaded at the same time. The collocated plan needs memory for only 8 * 125,000 = 1,000,000 rows at any one time. Collocated Hash Join Performance The collated join plan has disappointing performance in this case. The query runs for around 25,300ms despite the same IO statistics as usual. This is much the worst result so far, so what went wrong? It turns out that cardinality estimation for the single partition scans of table T1 is slightly low. The properties of the Clustered Index Scan of T1 (graphic immediately above) show the estimation was for 121,951 rows. This is a small shortfall compared with the 125,000 rows actually encountered, but it was enough to cause the hash join to spill to physical tempdb: A level 1 spill doesn’t sound too bad, until you realize that the spill to tempdb probably occurs for each of the 41 partitions. As a side note, the cardinality estimation error is a little surprising because the system tables accurately show there are 125,000 rows in every partition of T1. Unfortunately, the optimizer uses regular column and index statistics to derive cardinality estimates here rather than system table information (e.g. sys.partitions). Collocated Merge Join We will never know how well the collocated parallel hash join plan might have worked without the cardinality estimation error (and the resulting 41 spills to tempdb) but we do know: Merge join does not require a memory grant; and Merge join was the optimizer’s preferred join option for a single partition join Putting this all together, what we would really like to see is the same collocated join strategy, but using merge join instead of hash join. Unfortunately, the current query optimizer cannot produce a collocated merge join; it only knows how to do collocated hash join. So where does this leave us? CROSS APPLY sys.partitions We can try to write our own collocated join query. We can use sys.partitions to find the partition numbers, and CROSS APPLY to get a count per partition, with a final step to sum the partial counts. The following query implements this idea: SELECT row_count = SUM(Subtotals.cnt) FROM ( -- Partition numbers SELECT p.partition_number FROM sys.partitions AS p WHERE p.[object_id] = OBJECT_ID(N'T1', N'U') AND p.index_id = 1 ) AS P CROSS APPLY ( -- Count per collocated join SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals; The estimated plan is: The cardinality estimates aren’t all that good here, especially the estimate for the scan of the system table underlying the sys.partitions view. Nevertheless, the plan shape is heading toward where we would like to be. Each partition number from the system table results in a per-partition scan of T1 and T2, a one-to-many Merge Join, and a Stream Aggregate to compute the partial counts. The final Stream Aggregate just sums the partial counts. Execution time for this query is around 3,500ms, with the same IO statistics as always. This compares favourably with 5,000ms for the serial plan produced by the optimizer with the OPTION (MERGE JOIN) hint. This is another case of the sum of the parts being less than the whole – summing 41 partial counts from 41 single-partition merge joins is faster than a single merge join and count over all partitions. Even so, this single-threaded collocated merge join is not as quick as the original parallel hash join plan, which executed in 2,600ms. On the positive side, our collocated merge join uses only one logical processor and requires no memory grant. The parallel hash join plan used 16 threads and reserved 569 MB of memory:   Using a Temporary Table Our collocated merge join plan should benefit from parallelism. The reason parallelism is not being used is that the query references a system table. We can work around that by writing the partition numbers to a temporary table (or table variable): SET STATISTICS IO ON; DECLARE @s datetime2 = SYSUTCDATETIME();   CREATE TABLE #P ( partition_number integer PRIMARY KEY);   INSERT #P (partition_number) SELECT p.partition_number FROM sys.partitions AS p WHERE p.[object_id] = OBJECT_ID(N'T1', N'U') AND p.index_id = 1;   SELECT row_count = SUM(Subtotals.cnt) FROM #P AS p CROSS APPLY ( SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals;   DROP TABLE #P;   SELECT DATEDIFF(Millisecond, @s, SYSUTCDATETIME()); SET STATISTICS IO OFF; Using the temporary table adds a few logical reads, but the overall execution time is still around 3500ms, indistinguishable from the same query without the temporary table. The problem is that the query optimizer still doesn’t choose a parallel plan for this query, though the removal of the system table reference means that it could if it chose to: In fact the optimizer did enter the parallel plan phase of query optimization (running search 1 for a second time): Unfortunately, the parallel plan found seemed to be more expensive than the serial plan. This is a crazy result, caused by the optimizer’s cost model not reducing operator CPU costs on the inner side of a nested loops join. Don’t get me started on that, we’ll be here all night. In this plan, everything expensive happens on the inner side of a nested loops join. Without a CPU cost reduction to compensate for the added cost of exchange operators, candidate parallel plans always look more expensive to the optimizer than the equivalent serial plan. Parallel Collocated Merge Join We can produce the desired parallel plan using trace flag 8649 again: SELECT row_count = SUM(Subtotals.cnt) FROM #P AS p CROSS APPLY ( SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals OPTION (QUERYTRACEON 8649); The actual execution plan is: One difference between this plan and the collocated hash join plan is that a Repartition Streams exchange operator is used instead of Distribute Streams. The effect is similar, though not quite identical. The Repartition uses round-robin partitioning, meaning the next partition id is pushed to the next thread in sequence. The Distribute Streams exchange seen earlier used Demand partitioning, meaning the next partition id is pulled across the exchange by the next thread that is ready for more work. There are subtle performance implications for each partitioning option, but going into that would again take us too far off the main point of this post. Performance The important thing is the performance of this parallel collocated merge join – just 1350ms on a typical run. The list below shows all the alternatives from this post (all timings include creation, population, and deletion of the temporary table where appropriate) from quickest to slowest: Collocated parallel merge join: 1350ms Parallel hash join: 2600ms Collocated serial merge join: 3500ms Serial merge join: 5000ms Parallel merge join: 8400ms Collated parallel hash join: 25,300ms (hash spill per partition) The parallel collocated merge join requires no memory grant (aside from a paltry 1.2MB used for exchange buffers). This plan uses 16 threads at DOP 8; but 8 of those are (rather pointlessly) allocated to the parallel scan of the temporary table. These are minor concerns, but it turns out there is a way to address them if it bothers you. Parallel Collocated Merge Join with Demand Partitioning This final tweak replaces the temporary table with a hard-coded list of partition ids (dynamic SQL could be used to generate this query from sys.partitions): SELECT row_count = SUM(Subtotals.cnt) FROM ( VALUES (1),(2),(3),(4),(5),(6),(7),(8),(9),(10), (11),(12),(13),(14),(15),(16),(17),(18),(19),(20), (21),(22),(23),(24),(25),(26),(27),(28),(29),(30), (31),(32),(33),(34),(35),(36),(37),(38),(39),(40),(41) ) AS P (partition_number) CROSS APPLY ( SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals OPTION (QUERYTRACEON 8649); The actual execution plan is: The parallel collocated hash join plan is reproduced below for comparison: The manual rewrite has another advantage that has not been mentioned so far: the partial counts (per partition) can be computed earlier than the partial counts (per thread) in the optimizer’s collocated join plan. The earlier aggregation is performed by the extra Stream Aggregate under the nested loops join. The performance of the parallel collocated merge join is unchanged at around 1350ms. Final Words It is a shame that the current query optimizer does not consider a collocated merge join (Connect item closed as Won’t Fix). The example used in this post showed an improvement in execution time from 2600ms to 1350ms using a modestly-sized data set and limited parallelism. In addition, the memory requirement for the query was almost completely eliminated  – down from 569MB to 1.2MB. The problem with the parallel hash join selected by the optimizer is that it attempts to process the full data set all at once (albeit using eight threads). It requires a large memory grant to hold all 5 million rows from table T1 across the eight hash tables, and does not take advantage of the divide-and-conquer opportunity offered by the common partitioning. The great thing about the collocated join strategies is that each parallel thread works on a single partition from both tables, reading rows, performing the join, and computing a per-partition subtotal, before moving on to a new partition. From a thread’s point of view… If you have trouble visualizing what is happening from just looking at the parallel collocated merge join execution plan, let’s look at it again, but from the point of view of just one thread operating between the two Parallelism (exchange) operators. Our thread picks up a single partition id from the Distribute Streams exchange, and starts a merge join using ordered rows from partition 1 of table T1 and partition 1 of table T2. By definition, this is all happening on a single thread. As rows join, they are added to a (per-partition) count in the Stream Aggregate immediately above the Merge Join. Eventually, either T1 (partition 1) or T2 (partition 1) runs out of rows and the merge join stops. The per-partition count from the aggregate passes on through the Nested Loops join to another Stream Aggregate, which is maintaining a per-thread subtotal. Our same thread now picks up a new partition id from the exchange (say it gets id 9 this time). The count in the per-partition aggregate is reset to zero, and the processing of partition 9 of both tables proceeds just as it did for partition 1, and on the same thread. Each thread picks up a single partition id and processes all the data for that partition, completely independently from other threads working on other partitions. One thread might eventually process partitions (1, 9, 17, 25, 33, 41) while another is concurrently processing partitions (2, 10, 18, 26, 34) and so on for the other six threads at DOP 8. The point is that all 8 threads can execute independently and concurrently, continuing to process new partitions until the wider job (of which the thread has no knowledge!) is done. This divide-and-conquer technique can be much more efficient than simply splitting the entire workload across eight threads all at once. Related Reading Understanding and Using Parallelism in SQL Server Parallel Execution Plans Suck © 2013 Paul White – All Rights Reserved Twitter: @SQL_Kiwi

  Read the article

< Previous Page | 205 206 207 208 209 210 211 212 213 214 215 216  | Next Page >