Search Results

Search found 24891 results on 996 pages for 'update grub'.

Page 149/996 | < Previous Page | 145 146 147 148 149 150 151 152 153 154 155 156  | Next Page >

  • Why isn't Grub2 using custom resolution?

    - by beanaroo
    I hope somebody can help me with this... Running Ubuntu 12.10 I am trying to get Grub2 to use a resolution of 1600x900... I have checked vbeinfo and hwinfo --framebuffer which both provide: Mode 0x037f: 1600x900 (+6400), 24 bits This resolution has worked on previous Ubuntu and Fedora installs. Here is my /etc/default/grub: http://paste.ubuntu.com/1308444/ Line 25 = GRUB_GFXMODE="1600x900" As well as my /boot/grub/grub.cfg: http://paste.ubuntu.com/1308446/ Line 72 = set gfxmode=1600x900 I have tried using grub-customizer: Why is it not taking effect?? I have tried different resolutions... Backgrounds and font colours are changeable. Thank you for your insight.

    Read the article

  • Unknown filesystem error

    - by Ben
    My computer has (had) a dual boot of Ubuntu Precise and Windows 7. Recently, when I started the computer, grub gave me an error saying "Unknown filesystem", and sent me to grub rescue. I searched for solutions to this, and found this answer: Unknown filesystem error: grub rescue. I followed the instructions there, but they didn't really help: when I restarted the computer, it went straight to Windows without grub intervening at all (so it boots successfully, just not to the OS I want). This is the log from boot-repair. My Ubuntu partition is sda6. Naturally, I'm a little worried, since the log says "unknown filesystem type ''", which to me looks like the partition might be corrupt somehow. From the Ubuntu Live CD, I started gparted and looked at my partitions, and it also told me that my sda6 is an unknown filesystem. So my questions are basically: Can I restore Ubuntu to the way it was? If not, can I at least rescue the data I had on the partition somehow? Again, if I can't restore Ubuntu, how do I clean everything up so that I can reinstall it without too many complications? Another thing I should mention is that yesterday I had a similar problem where the computer told me there was a problem with the hard drive when it was starting, but it fixed itself by running fsck (that time it got past grub, and managed to start Ubuntu). Between that occasion and me having trouble booting the computer today, I have booted it successfully at least twice. Thanks for any help!

    Read the article

  • Why does Eclipse keep crashing and dpkg erroring out?

    - by Sanju Sony Kurian
    I AM GETTING THE ERROR E: Sub-process /usr/bin/dpkg returned an error code (1) WHILE I WAS TRYING TO UPGRADE MY PC. i USE ECLIPSE AND IT KEEPS ON CRASHING... pLEASE HELP I ATTACH THE ERROR sanju@sanju-Dell-System-XPS-L502X:~$ upgrade Reading package lists... Done Building dependency tree Reading state information... Done The following packages have been kept back: aisleriot gir1.2-rb-3.0 gir1.2-totem-1.0 gnome-disk-utility gnome-keyring libgck-1-0 libgcr-3-1 libtotem0 linux-generic linux-headers-generic linux-image-generic rhythmbox rhythmbox-data rhythmbox-mozilla rhythmbox-plugin-cdrecorder rhythmbox-plugin-magnatune rhythmbox-plugin-zeitgeist rhythmbox-plugins seahorse totem totem-common totem-mozilla totem-plugins 0 upgraded, 0 newly installed, 0 to remove and 23 not upgraded. 1 not fully installed or removed. After this operation, 0 B of additional disk space will be used. Do you want to continue [Y/n]? Y Setting up grub-pc (1.99-21ubuntu3.1) ... /var/lib/dpkg/info/grub-pc.config: 35: /etc/default/grub: Syntax error: EOF in backquote substitution dpkg: error processing grub-pc (--configure): subprocess installed post-installation script returned error exit status 2 Errors were encountered while processing: grub-pc E: Sub-process /usr/bin/dpkg returned an error code (1)

    Read the article

  • Win7 can't be loaded after installing ubuntu 11.10

    - by Kirill Lykov
    I use MacBookPro8,2. First of all I installed MacOS, then WIn7 on sde4, then Ubuntu on sde5 and swop on sde5. After installing ubuntu win7 isn't loaded when I choose it in grub. Win boot writes the following message: file: \boot\BCD Status: 0xc0000225 An error occurred while attempting to load the boot configuration data What shall I do? Try to repair windows using CD(in this case I think I will damage grub) or I can somehow configure grub to make it work properly?

    Read the article

  • Boot from Ubuntu ISO on a hfsplus partition (macbook pro)

    - by user279771
    I would like to be able to boot from an ISO stored on an HFS+ partition (the main partition on my macbook pro). Here is what I've done so far: (writing in shorthand :D) grub> insmod hfs,hfsplus,loopback,part_gpt grub> loopback loop (hd0,gpt2)/location/to/img.io grub> configfile (loop)/boot/grub/loopback.cfg ... This does not work. tab-complete of the (loop) path does not work... However, this does work (tab-complete and all) if the iso comes from my ext3 partition. For particular reasons, I can't have the iso images on the ext3 partition, they need to be kept on the hfs+ partition. What should be done?

    Read the article

  • Boot-up hangs unless I manually select the kernel in the bootloader

    - by The Photon
    In my stock Ubuntu install (forgot which critter its named for, about one year old), I am able to boot up normally only if I manually select the kernel in the bootloader (Grub). If I step away and it boots from the default, the system will hang with the word "Ubuntu" on a graphics screen and a few white/red blinking dots beneath it. umb@digdug:~$ uname -a Linux digdug 2.6.32-42-generic #95-Ubuntu SMP Wed Jul 25 15:57:54 UTC 2012 i686 GNU/Linux Synaptics says I have 'grub-pc' version 1.98-1ubuntu-13 installed. My grub.cfg does have default="0" which I would expect to make the default be the first kernel shown in the selection screen. The system is a laptop with an i7 processor, and I have had trouble with some kernels not being able to boot at all in the past, and power management is not working perfectly, but I have not had problems booting the latest kernel (2.6.32-42) if I select it manually in Grub. Any idea what is going on here and how can I fix it so that I can re-boot unattended?

    Read the article

  • Ubuntu 13.04 on UEFI system with Windows Boot Manager as the main loader

    - by Mehrdad
    On my old laptop (legacy BIOS, MBR disk), this was perfectly possible to get working: I turn on the computer and see the Windows Boot Manager I use EasyBCD (or BootPart, or something else) to add an option to the BCD menu which allows me to boot into GRUB, and then into Ubuntu I can't figure how to do this on my new laptop (UEFI, GPT disk), whether in UEFI or legacy mode. Currently I've installed (and even booted!) Ubuntu on my laptop, but only with the help of an external GRUB (on a USB flash drive). How can I add GRUB as an option in the Windows Boot Manager on a UEFI laptop? (No, I don't want to change my primary boot loader. So no, I don't want to overwrite the Windows boot loader with GRUB.)

    Read the article

  • Seeking for a better solution to restrict access in GRUB2 menu

    - by LiveWireBT
    I just read that in certain situations you should also protect access to your GRUB2 menu by setting a password and may be refining acces by adding --unrestricted or --users as arguments to menuentries und submenus. I read the corresponding pages in the Ubuntu Community Documentation and the Arch Wiki. So, I created /etc/grub.d/01_security, stored usernames and passwords in there, made the file executable and ran update-grub. This is working as intended, every action in the menu prompts for username and password, but I also want to modify the automatically generated entries to either restrict them to certain users (via --users) or make them available for everyone, but not editable by everyone (via --unrestricted). I was able to find the proper lines in 10_linux and edit them accordingly, however I'd love to see an easier solution. Perhaps an option like GRUB_DISABLE_RECOVERY="true" or GRUB_DISABLE_OS_PROBER=true in /etc/default/grub for easy (re)configuration (for linux and os-prober generated entries). Here's a diff from my 13.10 installation: $ diff /etc/grub.d/10_linux /etc/grub.d/10_linux_bak 123c123 < echo "menuentry '$(echo "$title" | grub_quote)' ${CLASS} --unrestriced \$menuentry_id_option 'gnulinux-$version-$type-$boot_device_id' {" | sed "s/^$ --- > echo "menuentry '$(echo "$title" | grub_quote)' ${CLASS} \$menuentry_id_option 'gnulinux-$version-$type-$boot_device_id' {" | sed "s/^/$submenu_inde$ 125c125 < echo "menuentry '$(echo "$os" | grub_quote)' ${CLASS} --unrestricted \$menuentry_id_option 'gnulinux-simple-$boot_device_id' {" | sed "s/^/$submenu_$ --- > echo "menuentry '$(echo "$os" | grub_quote)' ${CLASS} \$menuentry_id_option 'gnulinux-simple-$boot_device_id' {" | sed "s/^/$submenu_indentation/" 323c323 < echo "submenu --unrestricted '$(gettext_printf "Advanced options for %s" "${OS}" | grub_quote)' \$menuentry_id_option 'gnulinux-advanced-$boot_device_$ --- > echo "submenu '$(gettext_printf "Advanced options for %s" "${OS}" | grub_quote)' \$menuentry_id_option 'gnulinux-advanced-$boot_device_id' {" tl;dr: I'd love the see a simple solution for GRUB2 entries that cannot be modified without a password or are limited to certain users. (Yes, GRUB_DISABLE_RECOVERY="true" is active.)

    Read the article

  • Proper method to update and draw from game loop?

    - by Lost_Soul
    Recently I've took up the challenge for myself to create a basic 2d side scrolling monster truck game for my little brother. Which seems easy enough in theory. After working with XNA it seems strange jumping into Java (which is what I plan to program it in). Inside my game class I created a private class called GameLoop that extends from Runnable, then in the overridden run() method I made a while loop that handles time and such and I implemented a targetFPS for drawing as well. The loop looks like this: @Override public void run() { long fpsTime = 0; gameStart = System.currentTimeMillis(); lastTime = System.currentTimeMillis(); while(game.isGameRunning()) { currentTime = System.currentTimeMillis(); long ellapsedTime = currentTime - lastTime; if(mouseState.leftIsDown) { que.add(new Dot(mouseState.getPosition())); } entities.addAll(que); game.updateGame(ellapsedTime); fpsTime += ellapsedTime; if(fpsTime >= (1000 / targetedFPS)) { game.drawGame(ellapsedTime); } lastTime = currentTime; } The problem I've ran into is adding of entities after a click. I made a class that has another private class that extends MouseListener and MouseMotionListener then on changes I have it set a few booleans to tell me if the mouse is pressed or not which seems to work great but when I add the entity it throws a CME (Concurrent Modification Exception) sometimes. I have all the entities stored in a LinkedList so later I tried adding a que linkedlist where I later add the que to the normal list in the update loop. I think this would work fine if it was just the update method in the gameloop but with the repaint() method (called inside game.drawGame() method) it throws the CME. The only other thing is that I'm currently drawing directly from the overridden paintComponent() method in a custom class that extends JPanel. Maybe there is a better way to go about this? As well as fix my CME? Thanks in advance!!!

    Read the article

  • Grub2 attempting to boot hd1 when it should boot hd0

    - by JoBu1324
    I'm attempting to perform a "normal" install on a USB3 SSD (I don't know if it is noteworthy, but I don't have a swap partition). The installation proceeds normally (I'm installing from a USB2 device I created using LiLi Boot, with a copy of Ubuntu 12.10 64bit that I downloaded directly from the source. The system I'm running Ubuntu on has had a more traditional installation of ubuntu running on it without issue (also 12.10), so I know that everything works A-OK when booting from a 7200RPM internal disk. There are a number of oddities that I've noticed so far, including graphics corruption, but the first and most pressing issue is that Grub2 refuses to recognize the correct hd. From /boot/grub/grub.cfg: if [ x$feature_default_font_path = xy ] ; then font=unicode else insmod part_msdos insmod ext2 set root='hd1,msdos1' if [ x$feature_platform_search_hint = xy ]; then search --no-floppy --fs-uuid --set=root --hint-bios=hd1,msdos1 --hint-efi=hd1,msdos1 --hint-baremetal=ahci1,msdos1 b58ee4f7-d41d-400a-b7b8-18bd1f0ae9d3 else search --no-floppy --fs-uuid --set=root b58ee4f7-d41d-400a-b7b8-18bd1f0ae9d3 fi font="/usr/share/grub/unicode.pf2" fi This is from a 100% fresh install of linux (first boot), which was installed while no hard drives were connected to the system, other than the USB2 LiLi drive. The system refuses to boot unless I change the hd1,msdos1 - hd0,msdos1 in the grub menu at boot, when it is the only disk device connected to the PC. What options are left for me to troubleshoot this issue? I've been racking my brains and taxing the internet trying to dig up something on this problem, but now I'd like to see if the Ubuntu community can rise to the challenge and help me fix this boot problem. This is the second time I've attempted this particular setup. The first time, after days of wasted time, I managed to get it to boot every other boot - i.e. every even boot it would boot into Ubuntu like it was happy; every odd boot it would boot into the BusyBox or Grub prompt. At one point it complained that it couldn't find /dev/disk/by-uuid/[the disk], which I found most perplexing, since the disk was there and booted before and after the occurrence (with intervention).

    Read the article

  • All files gone after running fsck. How can I recover my files?

    - by cinlung
    I am a newbie in Linux. So this is my story I installed Ubuntu server 10.04lts. It worked great for many months, until today i decided to run fsck on the system partition and although it warned me, I kept pressing yes and now it will only boot into grub prompt. So i read some article and tried grub reinstall. But before performing grub reinstall, i decided to run fsck again from Ubuntu 10.04 lts for desktop live CD. The fsck painfully passes, now my drive is recognized as ext4 system and I am able to mount it again. However, all i can see is just boot directory and lost&found. I tried to perform grub reinstalling by doing grup-install stuff, now my grub is still not loading right, my files are missing, and the weird thing is that the amount I found used by boot and lost n found is only 5gb and the amount used in he hdd is 8 gb. So my files must be somewhere in the hdd. Is there any sinple way maybe a windows tool or something yo recover my files? I only need to retrieve my database backup and everything else can go. I am freaking out here. Please help.

    Read the article

  • Xorg and Virtualbox break after installing Xen

    - by stueng
    I installed XEN using these instructions https://help.ubuntu.com/community/Xen Now when I rebooted virtualbox doesnt work, X doesnt work, lightdm doesnt work How can I set it back to how it was before? I presume I want to revert the following actions: Modify GRUB to default to booting Xen: sudo sed -i 's/GRUB_DEFAULT=.*\+/GRUB_DEFAULT="Xen 4.1-amd64"/' /etc/default/grub sudo update-grub Can someone tell me how?

    Read the article

  • boot loading problem after wubi installation of Ubuntu 12.04 on Win7

    - by user63085
    I was new in Ubuntu OS and I was just trying to use wubi windows installer to get a Ubuntu first for hands-on. I followed exactly the same as the instructions and after reboot win7, there is no Ubuntu selection in windows boot manager, with only Windows 7 showing there evilly -.- What I've found out was that the grub folder inside Ubuntu folder ( in my C:\ drive) was empty, either inside the ubuntu\disks\grub or ubuntu\install\grub. I thought this might be the reason why I could not load ubuntu during startup. Cause I've also looked into the EasyBCD settings, and ubuntu entry with Bootloader Path: \ubuntu\winboot\wubildr.mbr was lying there peacefully, looking perfectly fine. However it was not in boot loader actually. Is there a way to restore the grub folder with grub2, or is there any way to fix this problem so that I can find the "Ubuntu" selection at windows startup? Very appreciate your help :) Henry

    Read the article

  • Errors when installing updates

    - by user71613
    I am getting the following errors when installing updates. They started to appear after I upgraded my system to 12.04. Errors were encountered while processing: samba-common samba-common-bin samba grub-pc grub-gfxpayload-lists Setting up samba-common (2:3.6.3-2ubuntu2.2) ... perl: error while loading shared libraries: libperl.so.5.12: cannot open shared object file: No such file or directory dpkg: error processing samba-common (--configure): subprocess installed post-installation script returned error exit status 127 dpkg: dependency problems prevent configuration of samba-common-bin: samba-common-bin depends on samba-common (>= 2:3.4.0~pre1-2); however: Package samba-common is not configured yet. dpkg: error processing samba-common-bin (--configure): dependency problems - leaving unconfigured dpkg: dependency problems prevent configuration of samba: samba depends on samba-common (= 2:3.6.3-2ubuntu2.2); however: Package samba-common is not configured yet. samba depends on samba-common-bin; however: Package samba-common-bin is not configured yet. dpkg: error processing samba (--configure): dependency problems - leaving unconfigured Setting up grub-gfxpayload-lists (0.6) ... Setting up grub-pc (1.99-21ubuntu3.1) ... perl: error while loading shared libraries: libperl.so.5.12: cannot open shared object file: No such file or directory dpkg: error processing grub-pc (--configure): subprocess installed post-installation script returned error exit status 127 Any ideas how to fix this?

    Read the article

  • windows system (bootloader) partition accidently deleted during multiple installs

    - by S.Y.T.
    After experimenting with multiple variations of backtrack and xbmcbuntu variations of Ubuntu with dual boot successfully, my windows partition became unrecognizable to grub. I used my windows boot CD to try to correct the problem. However, I deleted all partitions except for the NFTS one that contains my old windows install. (And, merged all other ones into that in hopes of getting back to the windows boot loader and out of grub) Now, all I get is a grub command prompt when I try and boot the system (how??? - I thought I deleted grub) And, now the windows boot disc doesn't even recognize the install. I've tried TRK to try and resolve the problem. Though I must admit ignorance in correctly using this utility. I've searched for other answers to this problem. Any help would be much appreciated. S.Y.

    Read the article

  • Why did Windows XP disappear from the menu on my triple-boot system?

    - by AS Chandrashekaran
    I have WIN XP installed, followed by Mint 12 and finally Ubuntu 12.04. I did a sudo grub-install /dev/sda followed by sudo update-grub. On rebooting all was fine and the boot menu had Ubuntu 12.04, mem, windows, and finally Mint 12. All was fine till now. Today wanted to boot into windows BUT found win XP missing from Boot Menu. Tried sudo grub-install /dev/sda followed by sudo update-grub BUT win XP is NOT FOUND. Can someone help with what to do to restore WIN XP in Boot Menu, Please?

    Read the article

  • How do I upgrade from ubuntu 9.10 to 12.10 on my Acer Aspire 3000

    - by 770
    I had my Acer Aspire 3000 as a dual boot XP/ubuntu 9.10 a couple years ago. I recently blew the dust of it and wanted to upgrade to 7/Ubuntu 12.10 so I began by formatting the Ubuntu side of the partition and apparently damaged the mbr as I could only get black screen with the error message: GRUB loading. error: no such partition grub rescue I then slaved the hdd to my win7 desktop and formatted the entire drive, both sides of the partition then reinstalled it in the Acer and tried to install win7. Upon starting the Acer I got the same error message: GRUB loading. error: no such partition grub rescue I then tried to reinstall Ubuntu 9.10 as I have an Ubuntu produced installation cd. Same result. Next day I received a new battery I had ordered for the Acer. I plugged it and the power supply in and hit the power button just to see if I at least could charge the battery but to my surprise Ubuntu 9.10 began to install, so I let it and it did. Now the hard drive shows 58 gb and 2.5gb partitions neither of which is formatted NTFS for/by windows. I am guessing that the GRUB/mbr was repaired somehow by the Ubuntu reinstallation. My question, should you choose to accept it; How can I get to my goal of dual boot win7/Ubuntu 12.10. I am a beginner and don't know much about linux or the terminology. Thank you for your thoughts and help.

    Read the article

  • Win7 and Ubuntu refuse to coexist

    - by Jeremy
    I'll make this quick: I have an HP laptop with win7, I installed Ubuntu on a separate partition, and when I tried to boot win7 from grub I got the loading screen and no progress-ever. I did a /fixmbr with the windows recovery cd and got back windows, but wiped out grub and my access to Ubuntu. I reinstalled grub from the Ubuntu live usb ( I know I did this correctly) and now windows won't boot, again. I'm a linux noob at a loss. Your wisdom is greatly appreciated! Update in response to Scott Severance: your instructions say to determine the main partition on my computer. I'm not sure what this means... my windows partition is at sda2, my boot partition is at sda1, and my linux root partition is at sda7... Which is the "main" partition? UPDATE: I determined that you were probably referring to the linux root(/) partition, because this was the only partition for which I could follow your instructions without errors. Now, Windows is booting fine (thanks to /fixmbr), but even after the grub instructions there is no grub. It boots straight into windows.

    Read the article

  • My computter wont shutdown

    - by marijn
    If I try to shutdown my computer it simply takes me back to the login-screen. I looked around and found : Type in terminal: 1. sudo gedit /etc/default/grub 2. Find the line: GRUB_CMDLINE_LINUX_DEFAULT="quiet splash" 3. Change this to: GRUB_CMDLINE_LINUX_DEFAULT="quiet splash acpi=force" 4. Save the file and close the file. 5. Finally, in terminal: sudo update-grub but now there is a new problem. if I try step 5 it says: myname@ubuntu:~$ sudo update-grub [sudo] password for myname: /usr/sbin/grub-probe: error: cannot find a device for / (is /dev mounted?). myname@ubuntu:~$ I am new whit linux and dont know what to do.

    Read the article

  • Error : [0.8879153] kernel panic -not syncing VFS unable to mount fs unknow block (8.3)

    - by user43069
    i installed ubuntu using wubit inside the windows and started working on it then i got this error afer updating [0.8879153] kernel panic -not syncing VFS unable to mount fs unknow block (8.3) and i can't user rescue mode and it's give me another error no filesystem could mount root ..... i looked at grub folder and didn' find any file on disks/boot/grub/ so i tryed to user super grub to fix it but it didn' work and it keep giving me. boot/grub/stage1 not found i didn't edit anything from grub folder. any idea plz .

    Read the article

  • Howto prevent grubpc 2.02 to display with submenu?

    - by useful
    In the past, we could tweak grub-pc so that it simply shows all OSes/options as a simple list, slightly editing /etc/grub.d/10_linux. How can we do that in the 2.02~beta2-9 release supplied with Trusty Tahr 14.04? With several kernels installed (eg generic and lowlatency) submenu is boring with such a /etc/default/grub when I need to change OS: GRUB_DEFAULT=saved GRUB_SAVEDEFAULT=true GRUB_TIMEOUT=3 # GRUB_HIDDEN_TIMEOUT=0 GRUB_HIDDEN_TIMEOUT_QUIET=true GRUB_DISTRIBUTOR=`lsb_release -i -s 2> /dev/null || echo Debian` GRUB_CMDLINE_LINUX_DEFAULT="acpi_osi=Linux" GRUB_CMDLINE_LINUX="" Thank you

    Read the article

  • SQL SERVER – Curious Case of Disappearing Rows – ON UPDATE CASCADE and ON DELETE CASCADE – Part 1 of 2

    - by pinaldave
    Social media has created an Always Connected World for us. Recently I enrolled myself to learn new technologies as a student. I had decided to focus on learning and decided not to stay connected on the internet while I am in the learning session. On the second day of the event after the learning was over, I noticed lots of notification from my friend on my various social media handle. He had connected with me on Twitter, Facebook, Google+, LinkedIn, YouTube as well SMS, WhatsApp on the phone, Skype messages and not to forget with a few emails. I right away called him up. The problem was very unique – let us hear the problem in his own words. “Pinal – we are in big trouble we are not able to figure out what is going on. Our product details table is continuously loosing rows. Lots of rows have disappeared since morning and we are unable to find why the rows are getting deleted. We have made sure that there is no DELETE command executed on the table as well. The matter of the fact, we have removed every single place the code which is referencing the table. We have done so many crazy things out of desperation but no luck. The rows are continuously deleted in a random pattern. Do you think we have problems with intrusion or virus?” After describing the problems he had pasted few rants about why I was not available during the day. I think it will be not smart to post those exact words here (due to many reasons). Well, my immediate reaction was to get online with him. His problem was unique to him and his team was all out to fix the issue since morning. As he said he has done quite a lot out in desperation. I started asking questions from audit, policy management and profiling the data. Very soon I realize that I think this problem was not as advanced as it looked. There was no intrusion, SQL Injection or virus issue. Well, long story short first - It was a very simple issue of foreign key created with ON UPDATE CASCADE and ON DELETE CASCADE.  CASCADE allows deletions or updates of key values to cascade through the tables defined to have foreign key relationships that can be traced back to the table on which the modification is performed. ON DELETE CASCADE specifies that if an attempt is made to delete a row with a key referenced by foreign keys in existing rows in other tables, all rows containing those foreign keys are also deleted. ON UPDATE CASCADE specifies that if an attempt is made to update a key value in a row, where the key value is referenced by foreign keys in existing rows in other tables, all of the foreign key values are also updated to the new value specified for the key. (Reference: BOL) In simple words – due to ON DELETE CASCASE whenever is specified when the data from Table A is deleted and if it is referenced in another table using foreign key it will be deleted as well. In my friend’s case, they had two tables, Products and ProductDetails. They had created foreign key referential integrity of the product id between the table. Now the as fall was up they were updating their catalogue. When they were updating the catalogue they were deleting products which are no more available. As the changes were cascading the corresponding rows were also deleted from another table. This is CORRECT. The matter of the fact, there is no error or anything and SQL Server is behaving how it should be behaving. The problem was in the understanding and inappropriate implementations of business logic.  What they needed was Product Master Table, Current Product Catalogue, and Product Order Details History tables. However, they were using only two tables and without proper understanding the relation between them was build using foreign keys. If there were only two table, they should have used soft delete which will not actually delete the record but just hide it from the original product table. This workaround could have got them saved from cascading delete issues. I will be writing a detailed post on the design implications etc in my future post as in above three lines I cannot cover every issue related to designing and it is also not the scope of the blog post. More about designing in future blog posts. Once they learn their mistake, they were happy as there was no intrusion but trust me sometime we are our own enemy and this is a great example of it. In tomorrow’s blog post we will go over their code and workarounds. Feel free to share your opinions, experiences and comments. Reference: Pinal Dave (http://blog.sqlauthority.com) Filed under: PostADay, SQL, SQL Authority, SQL Query, SQL Server, SQL Tips and Tricks, T SQL, Technology

    Read the article

  • SQL Server Table Polling by Multiple Subscribers

    - by Daniel Hester
    Background Designing Stored Procedures that are safe for multiple subscribers (to call simultaneously) can be challenging.  For example let’s say that you want multiple worker processes to poll a shared work queue that’s encapsulated as a SQL Table. This is a common scenario and through experience you’ll find that you want to use Table Hints to prevent unwanted locking when performing simultaneous queries on the same table. There are three table hints to consider: NOLOCK, READPAST and UPDLOCK. Both NOLOCK and READPAST table hints allow you to SELECT from a table without placing a LOCK on that table. However, SELECTs with the READPAST hint will ignore any records that are locked due to being updated/inserted (or otherwise “dirty”), whereas a SELECT with NOLOCK ignores all locks including dirty reads. For the initial update of the flag (that marks the record as available for subscription) I don’t use the NOLOCK Table Hint because I want to be sensitive to the “active” records in the table and I want to exclude them.  I use an Update Lock (UPDLOCK) in conjunction with a WHERE clause that uses a sub-select with a READPAST Table Hint in order to explicitly lock the records I’m updating (UPDLOCK) but not place a lock on the table when selecting the records that I’m going to update (READPAST). UPDATES should be allowed to lock the rows affected because we’re probably changing a flag on a record so that it is not included in a SELECT from another subscriber. On the UPDATE statement we should explicitly use the UPDLOCK to guard against lock escalation. A SELECT to check for the next record(s) to process can result in a shared read lock being held by more than one subscriber polling the shared work queue (SQL table). It is expected that more than one worker process (or server) might try to process the same new record(s) at the same time. When each process then tries to obtain the update lock, none of them can because another process has a shared read lock in place. Thus without the UPDLOCK hint the result would be a lock escalation deadlock; however with the UPDLOCK hint this condition is mitigated against. Note that using the READPAST table hint requires that you also set the ISOLATION LEVEL of the transaction to be READ COMMITTED (rather than the default of SERIALIZABLE). Guidance In the Stored Procedure that returns records to the multiple subscribers: Perform the UPDATE first. Change the flag that makes the record available to subscribers.  Additionally, you may want to update a LastUpdated datetime field in order to be able to check for records that “got stuck” in an intermediate state or for other auditing purposes. In the UPDATE statement use the (UPDLOCK) Table Hint on the UPDATE statement to prevent lock escalation. In the UPDATE statement also use a WHERE Clause that uses a sub-select with a (READPAST) Table Hint to select the records that you’re going to update. In the UPDATE statement use the OUTPUT clause in conjunction with a Temporary Table to isolate the record(s) that you’ve just updated and intend to return to the subscriber. This is the fastest way to update the record(s) and to get the records’ identifiers within the same operation. Finally do a set-based SELECT on the main Table (using the Temporary Table to identify the records in the set) with either a READPAST or NOLOCK table hint.  Use NOLOCK if there are other processes (besides the multiple subscribers) that might be changing the data that you want to return to the multiple subscribers; or use READPAST if you're sure there are no other processes (besides the multiple subscribers) that might be updating column data in the table for other purposes (e.g. changes to a person’s last name).  NOLOCK is generally the better fit in this part of the scenario. See the following as an example: CREATE PROCEDURE [dbo].[usp_NewCustomersSelect] AS BEGIN -- OVERRIDE THE DEFAULT ISOLATION LEVEL SET TRANSACTION ISOLATION LEVEL READ COMMITTED -- SET NOCOUNT ON SET NOCOUNT ON -- DECLARE TEMP TABLE -- Note that this example uses CustomerId as an identifier; -- you could just use the Identity column Id if that’s all you need. DECLARE @CustomersTempTable TABLE ( CustomerId NVARCHAR(255) ) -- PERFORM UPDATE FIRST -- [Customers] is the name of the table -- [Id] is the Identity Column on the table -- [CustomerId] is the business document key used to identify the -- record globally, i.e. in other systems or across SQL tables -- [Status] is INT or BIT field (if the status is a binary state) -- [LastUpdated] is a datetime field used to record the time of the -- last update UPDATE [Customers] WITH (UPDLOCK) SET [Status] = 1, [LastUpdated] = GETDATE() OUTPUT [INSERTED].[CustomerId] INTO @CustomersTempTable WHERE ([Id] = (SELECT TOP 100 [Id] FROM [Customers] WITH (READPAST) WHERE ([Status] = 0) ORDER BY [Id] ASC)) -- PERFORM SELECT FROM ENTITY TABLE SELECT [C].[CustomerId], [C].[FirstName], [C].[LastName], [C].[Address1], [C].[Address2], [C].[City], [C].[State], [C].[Zip], [C].[ShippingMethod], [C].[Id] FROM [Customers] AS [C] WITH (NOLOCK), @CustomersTempTable AS [TEMP] WHERE ([C].[CustomerId] = [TEMP].[CustomerId]) END In a system that has been designed to have multiple status values for records that need to be processed in the Work Queue it is necessary to have a “Watch Dog” process by which “stale” records in intermediate states (such as “In Progress”) are detected, i.e. a [Status] of 0 = New or Unprocessed; a [Status] of 1 = In Progress; a [Status] of 2 = Processed; etc.. Thus, if you have a business rule that states that the application should only process new records if all of the old records have been processed successfully (or marked as an error), then it will be necessary to build a monitoring process to detect stalled or stale records in the Work Queue, hence the use of the LastUpdated column in the example above. The Status field along with the LastUpdated field can be used as the criteria to detect stalled / stale records. It is possible to put this watchdog logic into the stored procedure above, but I would recommend making it a separate monitoring function. In writing the stored procedure that checks for stale records I would recommend using the same kind of lock semantics as suggested above. The example below looks for records that have been in the “In Progress” state ([Status] = 1) for greater than 60 seconds: CREATE PROCEDURE [dbo].[usp_NewCustomersWatchDog] AS BEGIN -- TO OVERRIDE THE DEFAULT ISOLATION LEVEL SET TRANSACTION ISOLATION LEVEL READ COMMITTED -- SET NOCOUNT ON SET NOCOUNT ON DECLARE @MaxWait int; SET @MaxWait = 60 IF EXISTS (SELECT 1 FROM [dbo].[Customers] WITH (READPAST) WHERE ([Status] = 1) AND (DATEDIFF(s, [LastUpdated], GETDATE()) > @MaxWait)) BEGIN SELECT 1 AS [IsWatchDogError] END ELSE BEGIN SELECT 0 AS [IsWatchDogError] END END Downloads The zip file below contains two SQL scripts: one to create a sample database with the above stored procedures and one to populate the sample database with 10,000 sample records.  I am very grateful to Red-Gate software for their excellent SQL Data Generator tool which enabled me to create these sample records in no time at all. References http://msdn.microsoft.com/en-us/library/ms187373.aspx http://www.techrepublic.com/article/using-nolock-and-readpast-table-hints-in-sql-server/6185492 http://geekswithblogs.net/gwiele/archive/2004/11/25/15974.aspx http://grounding.co.za/blogs/romiko/archive/2009/03/09/biztalk-sql-receive-location-deadlocks-dirty-reads-and-isolation-levels.aspx

    Read the article

  • Turned off my computer and o it's opened up in grub. How do I tell it to boot into lubuntu?

    - by Magpie
    It's just a commandline. I have tried my usual trick of sticking a usb with lubuntu in there to see if I can boot from my usb but I get a boot error (I normally wouldn't using that usb). GNU grub version 2.00-7ubuntu11 also just using basic things I know I hits ls then tried boot hd0 but it says: you need to load the kernel first I have no idea how to do this. I looked at the commands and most of them hve obscure names that are not helpful. UPDATE: I hit exit and that allowed be to boot from usb. Hopefully I will be able to find what I need to install to get it booting normally again from there.

    Read the article

  • Oracle HRMS APIs

    - by PRajkumar
    Oracle HRMS APIs..... Here I will be sharing all the Oracle HRMS APIs related articles. Item Type Author 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Create Employee API Update Employee API Create Employee Contact API Create Employee Address API Update Employee Address API Create Element Entries for Employee API Delete Element Entries for Employee API Rehire Employee API Create Employee Payment Method API Create and Update Employee Phone API Create and Update Employee Salary Proposal API Update Employee Fed Tax Rule API Create Employee State Tax Rule API Update Employee State Tax Rule API Update Employee Assignment API Puneet Rajkumar Puneet Rajkumar Puneet Rajkumar Puneet Rajkumar Puneet Rajkumar Puneet Rajkumar Puneet Rajkumar Puneet Rajkumar Puneet Rajkumar Puneet Rajkumar Puneet Rajkumar Puneet Rajkumar Puneet Rajkumar Puneet Rajkumar Puneet Rajkumar

    Read the article

< Previous Page | 145 146 147 148 149 150 151 152 153 154 155 156  | Next Page >