Search Results

Search found 46012 results on 1841 pages for 'unresponsive system'.

Page 501/1841 | < Previous Page | 497 498 499 500 501 502 503 504 505 506 507 508  | Next Page >

  • windows 7 turns off itself everyday at about 9am

    - by Radek
    I was given this comp at work and after a week or so this strange thing started to happen in the middle of doing something :-( It turns off itself at about 9.10am every morning Just once a day and it works fine after it had it little nap. it happens both if the comp was on all night or I turned it off before leaving the office. I tried to swapped the memory as I was told that there was an issue with memory. But moving or removing any memory did not make any difference. I am not aware that I would install any program that could cause that. I installed AVG and set it up to do every day scan about 8am + if restart is needed it requires user confirmation. 'The Software Protection service has stopped' few minutes before it turned off itself but it happened also at other times without the computer turned off. I turned off Windows automatic updates as the first thing when I got the comp configuration Windows 7 Ultimate Intel Core2 Quad Q9550 at 2.8GH, 8GB RAM ST31000528AS Barracuda 7200.12 SATA 3Gb/s 1TB Update Below message are logged when I turn the comp on again. Message 1: The previous system shutdown at 9:08:54 AM on ?6/?29/?2010 was unexpected. Message 2: Level:Critical Source: Kernel Power EventID 41 Task Category (63) The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly. Log after manual restart Update 2 (task scheduler)

    Read the article

  • How do you avoid that server documentation gets out of sync with the actual setup?

    - by Frerich Raabe
    I'm a hobbyist maintaining a small FreeBSD server serving mail via IMAP - it's an exercise in server administration. The setup does have reasonably good documentation (in AsciiDoc format) which recently allowed another person to recreate the entire setup from scratch in less than 30 minutes. However, I noticed that after the initial setup, it easily happens that small changes done to the system (say: inetd gets disabbled, my IMAP server listens on an additional port for ManageSieve connections, a new router is added to the exim configuration) don't end up in the documentation immediately (if at all). My idea was to avoid this problem by (partially?) generating the documentation out of the configuration files and the comments therein - one way to implement this may be to put /etc and /usr/local/etc into some source code management system (say - git) and then run a script which regenerates the documentation on every commit. However, I'm not sure whether that would be overkill and/or too difficult to get right (after all, I don't want complete copies of the source files in my documentation but rather just the diffs). How do other people avoid that the server documentation gets outdated - is there a good way to keep them in sync automatically, or do you just have the discipline to update the documentation the same time you modify the system?

    Read the article

  • Distributed storage and computing

    - by Tim van Elteren
    Dear Serverfault community, After researching a number of distributed file systems for deployment in a production environment with the main purpose of performing both batch and real-time distributed computing I've identified the following list as potential candidates, mainly on maturity, license and support: Ceph Lustre GlusterFS HDFS FhGFS MooseFS XtreemFS The key properties that our system should exhibit: an open source, liberally licensed, yet production ready, e.g. a mature, reliable, community and commercially supported solution; ability to run on commodity hardware, preferably be designed for it; provide high availability of the data with the most focus on reads; high scalability, so operation over multiple data centres, possibly on a global scale; removal of single points of failure with the use of replication and distribution of (meta-)data, e.g. provide fault-tolerance. The sensitivity points that were identified, and resulted in the following questions, are: transparency to the processing layer / application with respect to data locality, e.g. know where data is physically located on a server level, mainly for resource allocation and fast processing, high performance, how can this be accomplished? Do you from experience know what solutions provide this transparency and to what extent? posix compliance, or conformance, is mentioned on the wiki pages of most of the above listed solutions. The question here mainly is, how relevant is support for the posix standard? Hadoop for example isn't posix compliant by design, what are the pro's and con's? what about the difference between synchronous and asynchronous opeartion of a distributed file system. Though a synchronous distributed file system has the preference because of reliability it also imposes certain limitations with respect to scalability. What would be, from your expertise, the way to go on this? I'm looking forward to your replies. Thanks in advance! :) With kind regards, Tim van Elteren

    Read the article

  • What is the quickest and safest way to test new software and revert all changes, if needed?

    - by calbar
    I'm looking for Windows software that will allow me to quickly create a "checkpoint", do whatever I might need to do to my computer - install programs/drivers/updates, create/delete personal files, reboot the system multiple times, open questionable attachments - and then revert the entire system back to when the checkpoint was created. Essentially I want Windows Restore Points that save my personal files and partitions, too. It sounds like disk imaging might be the ticket, but creating them is much too slow and the restore process too involved... I'm hoping to sacrifice full disaster recovery for speed. Creating a checkpoint should be as close to one-click as possible, and rolling back should be a matter of selecting a restore point and rebooting. Ding! I'm familiar with Sandboxie, True Image Home "Try and Decide", Returnil, and a number of other "virtual system" apps that actively "catch" changes and allow you to commit or reject them. I'm not interested in these for a number of reasons - I prefer the "cut and dry" restore point approach. Finally, I'll note that I've just recently become aware of Comodo Time Machine. It sounds absolutely perfect, however, a quick skim through the user forums show more than a few horror stories of corrupted, unbootable systems. Any positive personal experience with the software to suppress my superstitions, or suggestions for more established alternatives would be greatly appreciated - Comodo Time Machine seems relatively new. Thanks for your help!

    Read the article

  • High memory utilization with firebird + windows server 2008 r2

    - by chesterman
    i have a Windows Server 2008 R2 (64bit) running a 64bit installation of Firebird 2.1.4.18393_0 in a 4GB phisical server. After a while, the task manager show that all memory is used, but the sum of the memory of all process does not stack to the half of the memory. Unfortunally, it's start swapping. Using RAMMAP, i can see that my entire database file is mapped into the memory. This only occours in windows server 2008 r2 and windows 7 64 bit. i can use firebird 32 or 64bit installations, doesn't matter. How can i prevent this? Why this only occours in w2k8r2 and w7? tks in advance ** UPDATE Aparently, this occours by the use of all memory by the file system cache. The microsoft documentations explain that this WAS a issue in windows xp, 2k3, vista and 2k8, but it was solved in 7 and 2k8r2. also adds that this issue is more common in 64bit hosts. (http://support.microsoft.com/kb/976618) there are some tools (DynCache, setcache and the Get/SetSystemFileCacheSize system calls from windows API) that allows me to fix a upper limit to the memory usage by the fscache, but the documentation argues that i should not do this in w2k8r2 because it will severely impact on the overall system performance. anyway, i tried, the performance remained the same shit and the use of the page file remained, although there is now more the 1gb of free memory.

    Read the article

  • task blocked for more than

    - by Manuel Sopena Ballesteros
    I have a webserver with the configuration below: VMWare ESXi environemt CPanel installed CentOS release 6.5 (Final) 4 CPUs 2G RAM 2x VM disks 100G each LVM system My issue is I am getting kernel panic quite frequently. These is a list of some processes blocked I could see from the console: mysqld queueprocd httpd suphp vmtoolsd loop0 auditd this is my sar logs Linux 2.6.32-431.3.1.el6.x86_64 (test01) 08/22/2014 _x86_64_ (4 CPU) 12:00:01 AM CPU %user %nice %system %iowait %steal %idle 12:10:01 AM all 26.86 0.01 0.98 0.57 0.00 71.57 12:20:01 AM all 1.78 0.02 1.03 0.08 0.00 97.09 12:30:01 AM all 26.34 0.02 0.85 0.05 0.00 72.74 12:40:01 AM all 27.12 0.01 1.11 1.22 0.00 70.54 12:50:01 AM all 1.59 0.02 0.94 0.13 0.00 97.32 01:00:01 AM all 26.10 0.01 0.77 0.04 0.00 73.07 01:10:01 AM all 27.51 0.01 1.16 0.14 0.00 71.18 01:20:01 AM all 1.80 0.07 1.06 0.08 0.00 96.99 01:30:01 AM all 26.19 0.01 0.78 0.05 0.00 72.96 01:40:01 AM all 26.62 0.02 0.87 0.05 0.00 72.45 01:50:02 AM all 1.35 0.01 0.87 0.02 0.00 97.75 02:00:01 AM all 26.11 0.02 0.69 0.02 0.00 73.17 02:10:01 AM all 26.73 0.02 0.89 0.14 0.00 72.21 02:20:01 AM all 1.45 0.01 0.92 0.04 0.00 97.58 02:30:01 AM all 26.59 0.01 1.06 0.03 0.00 72.31 02:40:01 AM all 26.27 0.01 0.72 0.05 0.00 72.95 02:50:01 AM all 0.86 0.01 0.50 0.09 0.00 98.53 03:00:01 AM all 25.61 0.02 0.39 0.03 0.00 73.96 03:10:01 AM all 26.30 0.08 0.66 0.14 0.00 72.82 03:20:01 AM all 0.81 0.01 0.51 0.04 0.00 98.63 03:30:02 AM all 26.15 0.02 0.53 0.07 0.00 73.24 03:40:01 AM all 26.06 0.01 0.47 0.04 0.00 73.42 03:50:01 AM all 0.96 0.02 0.51 0.03 0.00 98.48 Average: all 17.69 0.02 0.79 0.14 0.00 81.36 06:58:14 AM LINUX RESTART 07:00:01 AM CPU %user %nice %system %iowait %steal %idle 07:10:01 AM all 1.04 0.02 0.57 0.95 0.00 97.42 07:20:02 AM all 0.66 0.01 0.39 0.06 0.00 98.87 07:30:01 AM all 25.71 0.01 0.45 0.16 0.00 73.67 07:40:01 AM all 25.88 0.01 0.35 0.08 0.00 73.68 As you can see the server became unresponsive at 03.50 AM and I had to reset the VM at 06.58 AM to fix it. dmesg does not show any relevant information. I don't see any bottleneck in sar, any idea what can I check next? thank you very much

    Read the article

  • Windows Server 2008 - RAID 5 Fails on Reboot

    - by Adam
    Hey, I've got an install of Windows Server 2008 Enterprise. It's running software RAID-5 with five disks. The disks were originally formatted under Windows Server 2003, but came up fine once I installed Windows Server 2008. The issue I'm having is that every time I reboot the server, the RAID comes up with a "Failed Redundancy" - the data stays available. I have 4 disks on a PCI SATA controller, and one of the disks connected to the motherboard's on-board SATA ports. (The other on-board port has the system disk connected.) I was having Disk #4 fail consistently, so I tried swapping the cables on the controller end. I swapped the on-board RAID disk with one on the PCI controller. Same issue now, expect with disk #1. Once the system's up, I can reactivate the RAID, it will resync for a while, then go to "Healthy", and will stay that way for an indefinite amount of time - until I reboot. As soon as I reboot, the disk drops again. I've ruled out disk + cable with the recabling. I don't believe it would be the controller as it seems to work fine most of the time - only failing on reboot, and the other port on the same controller connects the system disk - which is clearly working. I did look in the event log, but didn't see anything particularly relevant (although I didn't know what I was looking for - just looked for anything with a "Warning" or "Error" symbol that looked disk-related :)). I'm not particularly familiar with RAID on Windows, does anyone have any idea why this might be doing this? Any idea how to fix it? Any suggestions appreciated! -- Adam

    Read the article

  • How do I log back into a Windows Server 2003 guest OS after Hyper-V integration services installs and breaks my domain logins?

    - by Warren P
    After installing Hyper-V integration services, I appear to have a problem with logging in to my Windows Server 2003 virtual machine. Incorrect passwords and logins give the usual error message, but a correct login/password gives me this message: Windows cannot connect to the domain, either because the domain controller is down or otherwise unavailable, or because your computer account was not found. Please try again later. If this message continues to appear, contact your system administrator for assistance. Nothing pleases me more than Microsoft telling me (the ersatz system administrator) to contact my system administrator for help, when I suspect that I'm hooped. The virtual machine has a valid network connection, and has decided to invalidate all my previous logins on this account, so I can't log in and remotely fix anything, and I can't remotely connect to it from outside either. This appears to be a catch 22. Unfortunately I don't know any non-domain local logins for this virtual machine, so I suspect I am basically hooped, or that I need ophcrack. is there any alternative to ophcrack? Second and related question; I used Disk2VHD to do the conversion, and I could log in fine several times, until after the Hyper-V integration services were installed, then suddenly this happens and I can't log in now - was there something I did wrong? I can't get networking working inside the VM BEFORE I install integration services, and at the very moment that integration services is being installed, I'm getting locked out like this. I probably should always know the local login of any machine I'm upgrading so I don't get stuck like this in the future.... great. Now I am reminded again of this.

    Read the article

  • Storing Windows Updates for reuse

    - by Saiyine
    At work we update lots of computers using Windows Update. Windows XP and 7 all day long, rarely some Vista. We do it through a corporate proxy, as connecting them to a domain to build a Wsus server is out of the question, so we download about two gigabytes a day of the very same updates everyday. I've tried WSUS Offline. It's pretty complete but when it finishes it's common to be still missing hundreds of megabytes of updates, because its intention is not to fully update a system but to install the critical updates, as the developers explain in the forums. Now I'm trying with Autoupdater. It's far worse, with poor capabilities for non-English Windows XP, but at least it gives the option to install non critical updates in Windows 7. It still misses hundres of megabytes of updates after fully updating the system. And finally, both doesn't install the driver related updates of Windows 7, so they at most save us a couple of hundreds of megabytes and a reboot (with the associated login to the computer and to the proxy) out of three or four. So, is it possible to somewhat extract the installed updates in a Windows 7 system and not having to download the same updates again and again at least with machines with the same hardware? Or even better, a generic package with all the updates?

    Read the article

  • User not in the sudoers file. This incident will be reported

    - by Sergiy Byelozyorov
    I need to install a package. For that I need root access. However the system says that I am not in sudoers file. When trying to edit one, it complains alike! How I am supposed to add myself to the sudoers file if I don't have the right to edit one? I have installed this system and only administrator. What can I do? Edit: I have tried visudo already. It requires me to be in sudoers in the first place. amarzaya@linux-debian-gnu:/$ sudo /usr/sbin/visudo We trust you have received the usual lecture from the local System Administrator. It usually boils down to these three things: #1) Respect the privacy of others. #2) Think before you type. #3) With great power comes great responsibility. [sudo] password for amarzaya: amarzaya is not in the sudoers file. This incident will be reported. amarzaya@linux-debian-gnu:/$

    Read the article

  • How to setup a virtual machine in Ubuntu desktop to run Debian Server

    - by stickman
    I want to run a virtual machine in my Ubuntu desktop that runs a Debian server. The purpose of this is to generate Debian packages. I have some C++ applications that were originally developed on my Ubuntu machine, and I need to (re)compile them on a Debian server in order to: build Deb packages for deployment on a Debian server make sure that the applications will definitely work on a debian server The idea is so that I can do 90% of my development on Ubuntu (where I am more comfortable), and deploy a binary package that definitely works on Debian. BTW, I am developing on Karmic Kola (Ubuntu 9.10). [Edit] Following the advice I got so far, I have installed debootstrap and Debian 'Lenny' on /srv/chroot/debian_lenny on my machine. I am not sure this is the server version, but in any case I dont think that matters for my purposes (though it would be useful to know how to specifically install the server version). At the moment though, I am like a fish out of water, since there is no GUI, and it is only a console that I have in the chroot jail. I had a look in the home folder (I cheated, by using the KNavigator in Ubuntu), and there are no folders there - which presumably mean that no users have been set up as yet in the Debian "system". I would like to know how to do the following: Download and install the dev tools needed for (re)compiling my C++ apps Copy my projects from the Ubuntu "system" to the Debian "system" After building the binaries, I would like to create a debian binary package containing all of my binaries, so that I can install the package on a Debian server (my remote server)

    Read the article

  • How to migrate WinXP from failing old HD to new one

    - by Péter Török
    Following this issue, we have all our important data backed up now. I also bought and installed a new replacement hard disk (WD 160GB PATA) as secondary (slave) drive. I created two primary NTFS partitions on it: a 40 GB system partition, and a 110GB data partition. In theory I could start reinstalling WinXP from scratch on the new system partition, then copying over all user data from the old drive to the new data partition. Once this is done, I could even throw away the old drive, or keep it just to see what happens. (Note: I don't want to clone the whole drive as it contains a dual boot setup with an old Linux installation which I don't need anymore, and anyway, a fresh reinstall would do WinXP good to get rid of many years' clutter.) However, I am lazy :-) The old HD is still functioning, the problem has not manifested again since. So I feel there is no need to hurry with a complete OS reinstall. What I don't know though is whether I will be able to install WinXP on the new system partition at a later stage without affecting the contents of the data partition on the same drive. If this is possible, I can just move over all our data to the new data partition to have it safe, then continue running WinXP from the old drive as long as it works. Does anyone see any problems/risks with this plan?

    Read the article

  • less maximum buffer size?

    - by Tyzoid
    I was messing around with my system and found a novel way to use up memory, but it seems that the less command only holds a limited amount of data before stopping/killing the command. To test, run (careful! uses lots of system memory very fast!) $ cat /dev/zero | less From my testing, it looks like the command is killed after less reaches 2.5 gigabytes of memory, but I can't find anything in the man page that suggests that it would limit it in such a way. In addition, I couldn't find any documentation via the google on the subject. Any light to this quite surprising discovery would be great! System Information: Quad core intel i7, 8gb ram. $ uname -a Linux Tyler-Work 3.13.0-32-generic #57-Ubuntu SMP Tue Jul 15 03:51:08 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux $ less --version less 458 (GNU regular expressions) Copyright (C) 1984-2012 Mark Nudelman less comes with NO WARRANTY, to the extent permitted by law. For information about the terms of redistribution, see the file named README in the less distribution. Homepage: http://www.greenwoodsoftware.com/less $ lsb_release -a No LSB modules are available. Distributor ID: Ubuntu Description: Ubuntu 14.04 LTS Release: 14.04 Codename: trusty

    Read the article

  • Is there a utility to visualise / isolate and watch application calls

    - by MyStream
    Note: I'm not sure what to search for so guidance on that may be just as valuable as an answer. I'm looking for a way to visually compare activity of two applications (in this case a webserver with php communicating with the system or mysql or network devices, etc) such that I can compare the performance at a glance. I know there are tools to generate data dumps from benchmarks for apache and some available for php for tracing that you can dump and analyse but what I'm looking for is something that can report performance metrics visually from data on calls (what called what, how long did it take, how much memory did it consume, how can that be represented visually in a call stack) and present it graphically as if it were a topology or layered visual with different elements of system calls occupying different layers. A typical visual may consist of (e.g. using swim diagrams as just one analogy): Network (details here relevant to network diagnostics) | ^ back out v | Linux (details here related to firewall/routing diagnostics) ^ back to network | | V ^ back to system Apache (details here related to web request) | | ^ response to V | apache PHP (etc) PHP---------->other accesses to php files/resources----- | ^ v | MySQL (total time) MySQL | ^ V | Each call listed + time + tables hit/record returned My aim would be to be able to 'inspect' a request/range of requests over a period of time to see what constituted the activity at that point in time and trace it from beginning to end as a diagnostic tool. Is there any such work in this direction? I realise it would be intensive on the server, but the intention is to benchmark and analyse processes against each other for both educational and professional reasons and a visual aid is a great eye-opener compared to raw statistics or dozens of discrete activity vs time graphs. It's hard to show the full cycle. Any pointers welcome. Thanks! FROM COMMENTS: > XHProf in conjunction with other programs such as Perconna toolkit > (percona.com/doc/percona-toolkit/2.0/pt-pmp.html) for mySQL run apache > with httpd -X & (Single threaded debug mode and background) then > attach with strace -> kcache grind

    Read the article

  • Looking for ballpark pricing on an affordable a Cisco VOIP solution for our office

    - by guytech
    We have about 8 incoming PSTN lines that are currently on an old and antiquated Nortel Meridian ICS system. This system has been giving us some grief. We're looking for a new VOIP solution. I've been looking at a Cisco solution and it does seem pricey but I'm sure effective. Unfortunately, we probably can't afford a Cisco Unified Communications 520 which seems to be the ideal solution. We have about 15 people who need an extension and voicemail. We really don't have any need for a fancy system just an auto attendant of some sort when people call us. It looks like we'll have to get an older router and an addon card for what we're looking for to get best value pricing. However, I don't know a a lot about Cisco voice products so I'm a bit lost as to what to get. The only thing I am sure on is the pricing on VOIP phones which we expect to be about ~$100-200. However, I'm not sure what pieces of VOIP infrastructure to get. Any advice? I am familiar with Asterisk but right now I'm looking on pricing concerning a Cisco solution.

    Read the article

  • Can a non-redundant RAID5 cause any serious problems (compared to RAID0)?

    - by leemes
    I used to have a three-disc RAID5 (mdadm) in my computer for personal media storage (music, videos, photos, programs, games, ...). It had three discs with 750 GB each, resulting in an array capacity of 1.5 TB. One day (one year ago), I needed one of those discs to install another operating system. I thought, I don't need the redundancy anymore since I backup the most important stuff (personal photos e.g.) on an external disc anyway. So I decided to remove one of the three discs without converting the RAID to RAID0 or even two separate discs, because I had no temporary storage (since one cannot simply convert the RAID5 to RAID0 AFAIK). So now, for about one year, I have a non-redundant RAID5 with 2 of 3 discs running. Sometimes, one of the discs has a defective contact at the power cable or something similar causing the drive to stop working temporarily (I don't know exactly what it is). Since it still works when rebooting the computer and in most cases by calling some mdadm commands, it wasn't that problematic. Note that the data is not very critical, since I still have a backup of the most important stuff. But in the last few weeks, one of the drives fails very frequently (every few hours), so it gets really annoying to manage this. My questions are: Is there any disadvantage (apart from the annoying management) of a non-redundant RAID5 (with one drive less than typical) over a RAID0? If I understand it correctly, both have no redundancy and the same capacity. On a temporary drive failure, I can restart the array in both cases, assuming that the drive itself still works after the failure. Can it happen that the drive contents alter on a drive failure, making the array inconsistent? If so, can I tell mdadm to check the array for failures (without a file system level checking tool)? Since the drive most probably only has a defective contact causing it to fail for a second only, can I tell mdadm to automatically restart the array, so I will not even notice the failure if no application wanted to access the file system during the failure?

    Read the article

  • How to access original files from before a symlink gets updated, which have since been moved to another dir

    - by Luke Cousins
    We have a website and our deployment process goes somewhat like the following (with lots of irrelevant steps excluded) echo "Remove previous, if it exists, we don't need that anymore" rm -rf /home/[XXX]/php_code/previous echo "Create the current dir if it doesn't exist (just in case this is the first deploy to this server)" mkdir -p /home/[XXX]/php_code/current echo "Create the var_www dir if it doesn't exist (just in case this is the first deploy to this server)" mkdir -p /home/[XXX]/var_www echo "Copy current to previous so we can use temporarily" cp -R /home/[XXX]/php_code/current/* /home/[XXX]/php_code/previous/ echo "Atomically swap the symbolic link to use previous instead of current" ln -s /home/[XXX]/php_code/previous /home/[XXX]/var_www/live_tmp && mv -Tf /home/[XXX]/var_www/live_tmp /home/[XXX]/var_www/live # Rsync latest code into the current dir, code not shown here echo "Atomically swap the symbolic link to use current instead of previous" ln -s /home/[XXX]/php_code/current /home/[XXX]/var_www/live_tmp && mv -Tf /home/[XXX]/var_www/live_tmp /home/[XXX]/var_www/live The problem we are having and would like help with is that, the first thing any website page load does is work out the base dir of the application and define it as a constant (we use PHP). If then during that page load a deployment occurs, the system tries to include() a file using the original full path and will get the new version of that file. We need it to get the old one from the old dir which has now moved as in: System starts page load and determines SYSTEM_ROOT_PATH constant to be /home/[XXX]/var_www/live or by using PHP's realpath() it could be /home/[XXX]/php_code/current. Symlink for /home/[XXX]/var_www/live get updated to point to /home/[XXX]/php_code/previous instead of /home/[XXX]/php_code/current where it did originally. System tries to load /home/[XXX]/var_www/live/something.php and gets /home/[XXX]/php_code/current/something.php instead of /home/[XXX]/php_code/previous/something.php I'm sorry if that is not explained very well. I'd really appreciate some ideas on how to get around this problem if someone can. Thank you.

    Read the article

  • The suggested way to handle pip(easy_install) with homebrew?

    - by Drake
    I know there are brew-gem and brew-pip but it is still really easy to get confused. Let's say my Mac OS X is 10.7.2. There are at least, as far as I know, 3 locations for Python modules (assume 2.7): /System/Library/Frameworks/Python.framework/Versions/2.7/ /Library/Python/2.7/site-packages /usr/local/lib/python2.7/site-packages/ (controlled within homebrew) For some Python modules, pip install them into 2, the so-called local/customized Python module location, and everything looks and works great. Ex, readline by *easy_install* (ipython suggested me to install readline by *easy_install* instead of pip) For some, it would try to install some miscellaneous files (ex, man, doc, ...) into system-wide location, which requires sudo! Ex, ipython insisted on installing man and doc into /System/Library/Frameworks/Python.framework/Versions/2.7/share/, which violates permission issue and all I can do is to use sudo. For some Python modules installed by brew, they are symbolic linked to /usr/local/lib/python2.7/site-packages/. Everything seems great except that you have to remember to add this location into PYTHONPATH. I am wondering any suggested and uniform way to handle those mass, or any explanation to make those stuff crystal clear.

    Read the article

  • Changing SkyDrive default location on Windows 8.1

    - by jmblack
    Update: This issue is due to the use of any non-local location for the SkyDrive folder. I've currently worked around this by using a local folder, but this has not resolved the original problem. I am trying to change the location where SkyDrive is located in Windows 8.1. This is the newly integrated SkyDrive, and not a downloadable client. There are a couple of similar questions on this site, but they pertain to Windows 8 and 8.1 Preview: Changing the SkyDrive default path in Windows 8.1 (Windows 8.1 Preview) How do I change the SkyDrive default folder in Windows (Windows 8) The SkyDrive folder is a system folder in 8.1, and like other system folders, has a Location tab in its properties that theoretically lets you change its location. I was able to do this with other system folders such as Downloads and Pictures, but there is an error when I try to change this for SkyDrive. Edit: The location that causes this error is a mapped network drive. The error does not occur when trying to move SkyDrive to another location on the same drive. There are no permission restrictions on the mapped network drive. Edit #2: This is the method I followed that produced the error: A screenshot of the error can be found here Has anyone else encountered this or have any information on this?

    Read the article

  • Computer not finding hard drives on boot -sometimes-

    - by todd.pund
    Computer specs: Mobo: Gigabyte ultradurable 3 - GA-970A-UD3 Processor: First gen I7 3.2GHZ Ram: 8GB Kingston DDR3 1066 Video Card: EVGA NVidia GTX 460 1GB Hard Drive: 500MB 7200rpm x2 (can't remember brand, sorry I'm at work.) Last week my developer preview for Windows 8 ran out so I put my copy of windows 7 back on the computer. The computer at that point started suffering from frequent freezing and crashing. When I rebooted the computer sometimes it wouldn't find the system HD at all. When I looked at the post screen it seemed to show that it wasn't finding either of the HDs. Then yesterday when turning on the computer I just got GRUB as a message (not a GRUB prompt, just GRUB) I haven't had a dual boot of Linux for at least a year. I loaded windows 7 recovery console from the disk and ran: bootrec /fixboot bootrec /fixmbr Which did not help. At that point I just installed Ubuntu 13.04 over the windows 7 install and still received the GRUB post. I went into the BIOS and switched the Hard Drive priorities and then it loaded into Ubuntu fine. For several days everything was just hunky dory until I installed the Ubuntu version of Steam, install Portal and tried to run it. At that point the computer froze and after hard rebooting couldn't find the hard disks again. Then after restarting the system it loaded up fine again and no issues since. (I have not tried to launch portal again). My next thought is to remove the system hard drive and try to use the secondary as the master to see if the primary HD is bad. I'm sorry if this has been confusing, I'll answer any questions I can. Any thoughts?

    Read the article

  • Update Grub on Squeeze - Kernel downgrade due VMware Server

    - by vodoo_boot
    I happen to run into various problems regarding grub and kernels. I don't really care about the kernel internas. All I want is VMware server in that dedicated root-server. 1.) What is a bzImage vs. vmlinuz? kaze:~# ls /boot/ System.map-2.6.32-5-amd64 bzImage-2.6.33.2 config-2.6.33.2 initrd.img-2.6.32-5-amd64 System.map-2.6.33.2 bzImage-2.6.35.6 config-2.6.35.6 vmlinuz-2.6.32-5-amd64 System.map-2.6.35.6 config-2.6.32-5-amd64 grub I updated my menu.lst (grub2): timeout 5 default 0 fallback 1 title 2.6.32.5 kernel (hd0,1)/boot/vmlinuz-2.6.32-5-amd64 root=/dev/sda2 panic=60 noapic acpi=off title 2.6.35.6 kernel (hd0,1)/boot//bzImage-2.6.35.6 root=/dev/sda2 panic=60 noapic acpi=off title 2.6.32.3 kernel (hd0,1)/boot//bzImage-2.6.33.2 root=/dev/sda2 panic=60 noapic acpi=off That doesn't do well... I think the vmlinuz file is missing initrd or so. Dunno. In fact I don't give too much about kernel boot voodoo as long as it works. update-grub(2) does not work. Does anybody know what magical trick there is to get the 2.6.32-5 booting? 2.) I thought t follow the Deban wiki.. I cannot get header-files for the installed 35.6 or 33.2 kernel in the repositories. I cannot build foreign headers because they will not match the running kernel. So how does one deal with that situtation? I'd prefer not to have to downgrade the kernel. Thanks for any answers!

    Read the article

  • PC won't boot from IDE HDD when SATA data drive connected

    - by Kevin
    I have an old Pentium 4 system running XP. The machine is set up as an HTPC. It was set up and running well with 1 SATA drive as a boot drive, another SATA drive to store TV recordings, and an IDE drive to store more recordings. Last week the original boot drive (a SATA drive) failed. The BIOS would no longer recognize it. I had a disused IDE drive hanging around that was large enough for the OS, so I reformatted it and installed XP on it. Now the system will only boot if I do not connect the remaining healthy SATA data drive. All three drives are recognized by the BIOS, and I have set the boot order so that the IDE drive with XP on it has top priority, but after the BIOS recognizes the drives, etc. I just get a black screen. I know the SATA drive is functional, because if I hot plug the drive AFTER the system is booted (I know I'm not supposed to do this), I can go into the control panels and mount the drive, and see all the files and folders on it in Windows Explorer. Any suggestions on what is going on and how to fix it? Many thanks.

    Read the article

  • umask seems to vary by user

    - by paullb
    I've got a development Ubuntu system for which I have several users: myself (with full sudo) and about 5 other users. (I've set up the system so everything in this respect is still at its default setting) I'm trying to set the system up so that multiple people can collaborate in a single directory by using grouing and I want the default permissions to be 664. However when some users edit files the permissions were 644. After a lot of investigating most users have a umask (checked at the prompt) of 0002 and when they create files they are 664 (as expected) but there are 2 (myself and one other) who have 0022 umask (so the files that come out are 644 and nobody else can write to them). I've looked everywhere but can't figure out why a couple users wind up with a different umask e.g. there is nothing the .bash_profile or anything like that) Any ideas for the source of the discrepancy? /etc/bashrc if [ $UID -gt 199 ] && [ "`id -gn`" = "`id -un`" ]; then umask 002 else umask 022 fi /etc/profile if [ $UID -gt 199 ] && [ "`id -gn`" = "`id -un`" ]; then umask 002 else umask 022 fi EDIT: My (bad) ~/.bashrc # .bashrc # Source global definitions if [ -f /etc/bashrc ]; then . /etc/bashrc fi # User specific aliases and functions export LANG=en_US.utf8 Other user (good) .bashrc # .bashrc # Source global definitions if [ -f /etc/bashrc ]; then . /etc/bashrc fi # User specific aliases and functions

    Read the article

  • Master File Table Corrupt, any way to save data?

    - by domen
    hi. I've used search, but none of the results match my problem so I didn't have to ask separate question. I've Installed Windows 7 RTM recently and since then partitions located on one of my HDDs have gone "crazy". They used to "freeze" and didn't open in explorer for some time (minute or two, usually), sometimes all partitions of the drive wouldn't show until reboot and finally, one of those partitions started showing "disk structure is corrupted and unreadable" warning, it appeared in Disk Management window as RAW and chkdsk showed "mft corrupt". There were no important data on the partition and I didn't have enough time to analyze the problem at the moment, so I just reformatted it and ran antivirus scan on system. After that problem settled for some time, but yesterday the problematic HDD vanished again from the system. After reboot chkdsk identified mft of four partitions corrupt and now they are all in same conditions as the above mentioned one. But the difference is that the files stored in them are extremely important. and just for info: I upgraded from Win7 build 7077, but had some performance issues, so I reformatted system drive and installed fresh Win7 RTM on it. I've downloaded TestDisk and it shows all the partitions marked as NTFS (not RAW) and my knowledge of the program wasn't sufficient to obtain any other info from it :-) and the images that could help describe the problem (sorry, I'm not allowed to post images and more than one hyperlink): http:// img22.imageshack.us/img22/5909/chkdskz.jpg http:// img198.imageshack.us/img198/5576/computeray.jpg I'm interested, is there a way to let me restore the MFT or just access files so I can backup them before reformatting the drive. Thanks for your time. :) P.S. my reformatted drive is showing no problems, could there be a problem with windows 7 itself? I googled, but with no results.

    Read the article

  • Use old raid drive as boot device without data loss

    - by Gabriel
    There were two disks in sw-raid. There were /dev/md1 as swap, /dev/md2 as boot and a /dev/md3 with ext4. The sw-raid was disabled by stopping and removing mdadm and then zeroing the superblock on each /dev/mdX partition with: sudo mdadm --zero-superblock /dev/sda1 sudo mdadm --zero-superblock /dev/sda2 sudo mdadm --zero-superblock /dev/sda3 In the disk that is the first boot device, I don't know if it's relevant, the system type of each partition was set back from fd to 82 or 83 with fdisk, /etc/fstab was updated, changing /dev/mdX to /dev/sdaX, and grub was reinstalled on the boot partition (/dev/sda2) with grub-instal. But the system wont boot. What else should I do to use this disk as the boot device without reinstall or data loss? Current output of fdisk Device Boot Start End Blocks Id System /dev/sda1 2048 33556480 16777216+ 82 Linux swap / Solaris /dev/sda2 * 33558528 34607104 524288+ 83 Linux /dev/sda3 34609152 3907027120 1936208984+ 83 Linux With it doesn't boot I mean that it stops in the grub console (with the grub> symbol). A ls command says: (hd0) (hd0,msdos3) (hd0,msdos2) (hd0,msdos1) (hd1) (hd1,msdos1) It's weird because hd1 was formatted with ext4...

    Read the article

< Previous Page | 497 498 499 500 501 502 503 504 505 506 507 508  | Next Page >