Search Results

Search found 5987 results on 240 pages for 'ghost recon future intern'.

Page 51/240 | < Previous Page | 47 48 49 50 51 52 53 54 55 56 57 58  | Next Page >

  • AWStats consumes too much resource, how to disable temporarily

    - by trante
    For some days AWStats takes %10-%20 of my CPU, takes 400-550 MB RAM and works for hours. Maybe my site's traffic became larger so process time takes more time than before or some bugs in program makes this. Anyway I want to disable AWStats temporarily. Maybe I would want to activate it in future. I found that answer. But it gives commands to remove AWStats. I only want to disable it temporarily. My system is Centos 6.3, Plesk 11.5.30 Update #19. I tried to disable cron jobs. I run this # killall awstats.pl I opened # vi /etc/cron.daily/awstats file and I changed it to this: #!/bin/sh #/usr/share/awstats/awstats_updateall.pl now -awstatsprog=/var/www/cgi-bin/awstats/awstats.pl -configdir=/etc/awstats >/dev/null 2>&1 exit 0 After some time I still see that awstats is running. What should I do more to not to awstats run again ? But without removing my files. After changing " /etc/cron.daily/awstats" file awstats doesn't start in daytime. But every night in 03:15 awstats starts again. Because of Plesk auto updates are working at that time, I changed from Plesk. Don't auto update automatically. But it seems like last night at 03:15 awstats started again. Is there any way to stop awstats temporarily except this solution ? Because this solution deletes awstats configs permanently and I don't know how to revert it back in future ? Turn off all AWStats for Plesk 11+ domains #!/bin/bash for i in /var/www/vhosts/*; do echo "Turning off and deleting Stats for" echo `basename $i` /usr/local/psa/admin/bin/webstatmng --unset-configs --stat-prog=awstats --domain-name=`basename $i` /usr/local/psa/admin/bin/webstatmng --clean --stat-prog=awstats --domain-name=`basename $i` done

    Read the article

  • Sony Vaio Hard Drive replacement won't boot

    - by AJ Bovaird
    I'm working on a Sony Vaio VGN-CR120e. I've cloned the original SATA 1.5 drive with Norton Ghost on to 2 new hard drives, the second of which is a Seagate 7200.4 320Gb 7200 RPM, with the jumper set to limit to SATA 1.5. (The first drive is a WD Scorpio Black 320Gb 7200RMP, which has no SATA 1.5 override support). Neither of the cloned drives will boot - I immediately get a BlackSOD saying: "Windows failed to start blah blah blah" "File \Windows\system32\winload.exe" "Status: 0xc000000e" "Info: The selected entry could not be loaded because the application is missing or corrupt." I've done this hundreds of times on other PC's, and this is the first time I've encountered such an error. Does anyone have any suggestions or advice on how to proceeed, as I would rather not reinstall Vista unless absolutely necessary.

    Read the article

  • Setting up SSL for phpMyAdmin

    - by Ubuntu User
    I would like to run phpmyadmin using my SSL certificate. I read that if I placed the following within the file: /etc/phpmyadmin/config.inc.php, it would force it to use SSL. And now it does... $cfg['ForceSSL'] =true; However, my issue is when I did this, now I get an error stating "cannot connect to server." I do a port scan and my port 443 is closed for one, but I am connecting via https:// for my secure web based email admin panel. This tells me this may not be the issue. Second, is that I have a SSL certificate I purchased but I am not sure how to apply this cert. mydomain.com.crt is sitting on my desktop, how should I be utilizing this? I remember creating a self signed cert for my web-email access. Do I have to do this for phpmyadmin as well? At least this way, since I am the only one who will ever access the DB, it will never expire. Also the phpmyadmin used to come up as: http://mydomain/phpmyadmin/ of course I am now trying to get to https://mydomain.com/phpmyadmin/ however, I do not have any pages on my website that requires https:// currently. In the future I may add this. But for now, I only want to access phpmyadmin via ssl. I can use my own -- but if this causes problems with future ecommerce apps under mydomain.com I would rather use the SSL cert I already purchased. Thank you!

    Read the article

  • Cisco QoS Guidence

    - by Kyle Brandt
    I have a 10M connection to the internet that is hooked into a 100M port. I am getting started with QoS, and am hopping for a little guidance on setting it up on a Cisco 3825 router. Right now I am going forward with the idea that I have to implement it on my router, and the provider can't provide QoS for me. How I envision it working is that the QoS will drop or queue packets on my router and that will help prevent a situation where the provider has to start dropping a lot of packets. Right now all I am tasked with is making sure that one of the 3 LANs gets a certain slice (say 3M for Gig Lan1) of the 10M internet connection (But ideally this will be more flexible in the Future). 10M Internet on 100M port on HWIC-4ESW +-----------------------+ | | Gig Lan1 | Cisco 3825 | Lan3 on HWIC-4ESW | | +-----------------------+ Gig Lan2 I need to learn more about QoS, but having a target technology and maybe example configuration will help me wrap my head around the reading I am doing a little more. Which Cisco QoS Technology do you recommend for this particular situation? Have a basic sample config of how this might work? Right now the 10M line is not congested, so this more to have something in place in case it starts to become mildly congested in the future.

    Read the article

  • Temporary boot problem after thunder storm - likely causes?

    - by alastairs
    The village where I live was sat under a thunder cloud for most of Friday, and we suffered a few power fluctuations (specifically, what seemed to be split-second outages). When I got back home from work, I found that my PCs had shut down during one of these outages. When I went to boot one of them back up, I couldn't get anything to display on screen, nor did the boot seem to complete correctly. I tried a number of things - unplugging different bits of hardware, swapping graphics adaptors, etc. - to no avail. I thought I was looking at a fried motherboard or CPU. Power seemed to be distributed correctly to the peripherals (the drives all appeared to be working) so I figured it couldn't be the PSU. Eventually I unplugged it from the mains and left it overnight (approx 12hrs unplugged). I tried it again this morning, and it booted up correctly. Woo-hoo! I have all my equipment protected by surge-protected power strips, so I don't think a spike caused these problems. Obviously it has something to do with the power fluctuations, and maybe the PSU in the problem machine got itself confused somehow. The questions are, for future reference and to help people with similar problems: What are the likely causes of the boot failure I experienced? Is a UPS a simple and cost-effective solution, or might other things help prevent this happening in future? What UPS can you recommend (my budget is limited)?

    Read the article

  • VoIP setup for one external PSTN line

    - by Jcl
    I'm completely new to VoIP and the likes, and I'm trying to find information about what could be the best setup for this. I need 4 (maybe more in the future, but maximum 5 or 6) wireless extensions, connected to 1 PSTN line, and maybe 2 in the future. I've been trying to gather information about the gear needed but everything I find seems too much over-the-top (and extremely expensive). The main problem is that the physical place we are on doesn't have possibilities of having a decent internet connection, so using a external VoIP "virtual PBX" is not an option. Thing is, even if small, phone is critical to this organization. I currently have an analog DECT/GAP PBX which does what I need, however the PBX is very bad and the call quality is horrible, and that's why I want to change it. The requirements would be: 4 wireless terminals (routing cable is not an option), all of them ringing on incoming PSTN calls. Ability to do internal calls (4 separate offices) and ability to pass calls between terminals. The 4 terminals should be able to access the external PSTN line without dialing any special codes. Very important: terminals should be able to issue commands on the PSTN line to the external operator in the form *nn*nnnnnnnn# . Don't know wether this could face to be a problem, but I've had problems with analog PBX which would take any * as a PBX command and wouldn't allow terminals to send it to the external lines. Not so important, but would be nice to have: call waiting music Could anyone recommend such a setup? I need to be able to do this on a EXTREMELY LIMITED budget (that is: I don't have a limit, but all should get as much to zero as possible). I have enough spare powerful computers and a 300mbps wireless network which works just fine, so that's not to include in the budget. Don't really know if this is the best place to ask, but it's the most StackExchange-related site I've found to this subject.

    Read the article

  • Create Google Maps screenshots at regular intervals

    - by Dave Jarvis
    Background People are concerned that building a pipeline to the West Coast of Canada will increase the number of oil tankers, thus increasing the probability of a major oil spill, thereby creating an environmental catastrophe. The AIS Live Ships Map website captures real-time Marine Traffic updates using a Google Maps interface. While it is possible to obtain data from an AIS data feed, often the feeds are either pay-for-use, or otherwise encumbered with license restrictions. Problem The AIS Live Ships website presents a map in the browser: The map above has had its location interactively changed to focus on the area in question: the northern straight of Vancouver Island. Question How would you create a service that captures the map every 30 minutes and that could run, with neither user-intervention nor a significant memory footprint, for a few years? Idea #1 Create a virtual machine. Install and run a light-weight browser. Use Shutter to take captures at regular intervals. Idea #2 Use Python's Ghost Webkit to automate the captures. Thank you!

    Read the article

  • Memory cache Ubuntu 9.10 server x86 doesn't work as expected

    - by Matthijs
    We're using an Ubuntu 9.10 server to transfer Ghost-image files. We configured it only with Samba. And the DOS-clients connect to Samba. The latest updates are installed and so far the servers is running fine. When we image 10 pc's with the same image of 2 files of 2GB there's no disk activity. Everything is loaded in the RAM. There's 4GB in the server. But when we use 2 pc's with 2 different image files of 500 MB (8x) files then there's a lot of continuous disk activity. The speed is lower. So it seems that Ubuntu doesn't cache more then one big file. Are there settings to change this behaviour?

    Read the article

  • Why is Apple System Image Utility so slow?

    - by Jon Rhoades
    I'm using Apple System Image Utility (SIU) on Snow Leopard 10.6.2 and I am rather disturbed it takes over Three hours to make a Netrestore or Netboot image. I'm using as the donor machine a brand new iMac and as the imaging machine a brand new iMac connected using target disk mode & Firewire 800. The hard drive size and subsequent image is about 8GB. To restore the image over the network takes about 4 minutes. Given that Norton Ghost will take an image in about 5 minutes (or less on newer machines) over USB2, why is the Mac over an order of magnitude slower?

    Read the article

  • Suggestions for Windows 8 migration [closed]

    - by Big Endian
    I'm thinking of migrating to Windows 8. At first I hated it, but I'm pretty sure the Windows 8 model is the future, and I don't particularly want to end up hating the future like my parents, frustrated and bewildered by anything past Windows XP. I'm currently running Windows 7 and my system has been accumulating some problems. It's probably an accumulation of issues from installing too much software, changing firewall settings, installing Ubuntu alongside Windows, and... well I'm not sure, but my computer has been buggy in unexpected ways lately (freezing and unfreezing, display driver crashing and recovering, and what I call "deep freeze/thaw cycle" where the mouse won't even move for a while). I'm good at solving computer problems, but I can't seem to get to the root of these and my best idea for fixing them is making sure I've backed up every file then re-installing the entire OS. Luckily for me, a new OS is just around the corner so this would be a good time to get two things out of the way at once. The problem I see is that the upgrade options I see are all "seamless". I don't want a seamless upgrade. I want to wipe the slate clean and start all over. Does this mean I will have to buy a full, new copy of Windows 8 rather than one of the cheaper upgrading options? Or does it not make since for me to go to Windows 8 given that I have a laptop, not a tablet? Maybe I should just re-install Windows 7, or even call good enough good enough, try to eliminate the bugs, and start with a fresh slate in 2-3 years after this computer eventually dies entirely from (inevitable) hardware failure. What would be the advantages or disadvantages and costs of each option, how would I go about upgrading to Windows 8 if that's the option I choose, and what is your personal opinion about my situation?

    Read the article

  • Updating a backup image (.wim and/or Acronis .tib)

    - by Backdraft
    Anyways, I've got a Windows 7 installation that I want to make a generalized backup image of so I can use it for future installs on not only my desktop from which the image is to be derived from, but also other systems with dissimilar hardware. Therefore I've arrived at either 2 options, using either sysprep/imagx from WAIK (guide here), or the simpler Acronis True Image w/ their Universal Restore addon. Of course, they create distinct image file types, .wim and .tib respectively. What I'd like to do is to periodically update this image, say with Windows Updates, by booting it to either a physical partition or using virtualization (VirtualBox/VMWare), perform the updates, and save the updated .wim or .tib image file again. What's the simplest way I could do this? Another question is, I created this generalized backup image on a 500GB Seagate 7200RPM HDD. Say I get an SSD as an OS drive in the future, can I just deploy this backup image to the SSD normally, or are there any potential problems to be aware/avoid (ie. is it best to completely reinstall the OS on the SSD from scratch, or can I use the image created on the normal HDD with no issue)? Thanks and Happy Holidays.

    Read the article

  • How to generate a round-numbers graph in Excel?

    - by tcheregati
    folks! Now, I have an Excel file with measurements I made of some color patches (I work at a Press company), with a device called spectrophotometer. Here it is: https://docs.google.com/open?id=0B0i8fdSf2ihzRlFYNWd4anItenM Density and Hue are two characteristics of each color patch. The thing is: I'm looking at a non-linear increase between the 25 Color Density measurements I took, but I NEED to know exactly how the color's Hue changes as the color's Density increases. For that, I needed Excel to give me round numbers for the X axis (for example 0,70 to 1,50 in 0,05 increments). And for that, obviously, I needed Excel to calculate the probable Hue Values corresponding to those ghost/round/not-given values of Density (like a kind of advanced rule of three). So, can anyone help me on that? Thanks a lot!

    Read the article

  • Cisco QoS Guidance

    - by Kyle Brandt
    I have a 10M connection to the internet that is hooked into a 100M port. I am getting started with QoS, and am hopping for a little guidance on setting it up on a Cisco 3825 router. Right now I am going forward with the idea that I have to implement it on my router, and the provider can't provide QoS for me. How I envision it working is that the QoS will drop or queue packets on my router and that will help prevent a situation where the provider has to start dropping a lot of packets. Right now all I am tasked with is making sure that one of the 3 LANs gets a certain slice (say 3M for Gig Lan1) of the 10M internet connection (But ideally this will be more flexible in the Future). 10M Internet on 100M port on HWIC-4ESW +-----------------------+ | | Gig Lan1 | Cisco 3825 | Lan3 on HWIC-4ESW | | +-----------------------+ Gig Lan2 I need to learn more about QoS, but having a target technology and maybe example configuration will help me wrap my head around the reading I am doing a little more. Which Cisco QoS Technology do you recommend for this particular situation? Have a basic sample config of how this might work? Right now the 10M line is not congested, so this more to have something in place in case it starts to become mildly congested in the future. I do have VOIP at one location connected to this one over the Internet that goes through a VPN tunnel. Everything else that is between this location and other offices is on a separate MPLS network.

    Read the article

  • Apache2 - rewrite a bunch of specified pathname URLs to one URL

    - by James Nine
    I need to rewrite a bunch of urls (about 100 or so) for SEO purposes, and there may be more being added in the future (probably another 50-100 later on). I need a flexible way of doing this and so far, the only way I can think of is to edit the .htaccess file using the rewrite engine. For example, I have a bunch of urls like this (please note that the query string is irrelevant, and dynamic; it could be anything. I was only using them purely as an example. I am only focusing on the pathname--the part between the hostname and query string, as marked in bold below): http://example.com/seo_term1?utm_source=google&utm_medium=cpc&utm_campaign=seo_term http://example.com/another_seo_term2?utm_source=facebook&utm_medium=cpc&utm_campaign=seo_term http://example.com/yet_another_seo_term3?utm_source=example_ad_network&utm_medium=cpc&utm_campaign=seo_term http://example.com/foobar_seo_term4 http://example.com/blah_seo_term5?test=1 etc... And they are all being rewritten to (for now): http://example.com/ What's the most efficient/effective way of doing this so that I may be able to add more terms in the future? One solution I came across is to do this (in the .htaccess file): RewriteEngine on RewriteCond %{REQUEST_FILENAME} !-f RewriteCond %{REQUEST_FILENAME} !-d RewriteRule ^(.*)$ / [NC,QSA] However, the problem with this solution is that even invalid urls (such as http://example.com/blah) will be rewritten to http://example.com instead of giving a 404 code (which is what it is supposed to do anyway). I'm still trying to figure out how all this works, and the only way I can think of is to write 100 more RewriteCond statements (such as: RewriteCond %{REQUEST_URI} =/seo_term1 [NC,OR]) before the RewriteRule directive. For example: RewriteEngine on RewriteCond %{REQUEST_FILENAME} !-f RewriteCond %{REQUEST_FILENAME} !-d RewriteCond %{REQUEST_URI} =/seo_term1 [NC,OR] RewriteCond %{REQUEST_URI} =/another_seo_term2 [NC,OR] RewriteCond %{REQUEST_URI} =/yet_another_seo_term3 [NC,OR] RewriteCond %{REQUEST_URI} =/foobar_seo_term4 [NC,OR] RewriteCond %{REQUEST_URI} =/blah_seo_term5 [NC] RewriteRule ^(.*)$ / [NC,QSA] But that doesn't sound very efficient to me. Is there a better way?

    Read the article

  • Small Business HP Virtualisation and iSCSI SAN Options

    - by Robin Day
    We are a small business that hosts our core product on a number of HP servers. Our core production setup is 1x HP DL380, high powered for a SQL Server Database 1x HP DL360, mid powered for our core application server 6x HP DL320, low powered for our front ends We run our training / testing / support systems on a similar setup, the servers are just older and less powerful. Unfortunately this is now causing us issues as the system has grown beyond the capabilities of these older servers. Upgrading these servers would be expensive and we believe that virtualisation is probably the way to go for the future. Locally we run a number of test / dev environments on ESXi using Direct Storoage on a couple of high powered DL360's and these are performing fairly well. We're thinking that instead of replacing all of our test servers that we can implement an iSCSI SAN and one or two high powered hosts. Hopefully looking that when it comes to replace our live servers as well that we can just expand the virual environment to cope. So my question is... Can anyone offer any advice on some suitable options? We have generally always been extremely happy with HP servers, all of our kit is currently HP, therefore our preference would be to stick with HP, however, I'm always happy to hear about other options. I'm hoping that initially a budget of around 15-25k (GBP) would be suitable, this could potentially be increased if I had confidence that the system would pave the way for a cost effective upgrade of our live systems in the future as well. I am new to SAN's and my only real experience is playing with OpenFiler on some old desktops. I think iSCSI should be suitable, but I've not done any research into how SQL server may perform. I've had a browser through HP's sites and see plenty of information about EVA, MSA, LeftHand, etc. However, from looking at all that, I don't see which options would be best and more importantly I don't know exactly what I would need to buy. Any help, links, opinions would be much appreciated. Thanks

    Read the article

  • How to copy partition from one disc to another (boot partition keeping all the vital data)?

    - by Patryk
    I have bought a new laptop but the HDD, which runs at 5400 rpm, is not sufficient for me. The laptop runs Windows 7 64-bit. I have my 'old' one (a better one - Seagate Momentus 7200 rpm) and I would like to replace it but without reinstalling everything. And there my question arises: can I copy my boot partition from my laptop hard drive to my old drive so that it will boot from it properly? If so, then how to do it? Will Norton Ghost be useful here? My point would be to just replace this partition and leave the rest.

    Read the article

  • NGINX server_name issues

    - by Unai
    I have the following simple server block on NGINX: server { listen 80; listen 8090; server_name domain.com; autoindex on; root /home/docroot; location ~ \.php$ { include /usr/local/nginx/conf/fastcgi_params; fastcgi_index index.php; fastcgi_param SCRIPT_FILENAME /home/docroot$fastcgi_script_name; fastcgi_pass 127.0.0.1:9000; } } After I include the relevant settings on my hosts file I get the following (unexpected) behavior: http: //domain.com/ and http: //domain.com:8090/ work fine; http: //domain.com:8090/future-cell-phone-technology-01-150x150.jpg works; http: //domain.com/future-cell-phone-technology-01-150x150.jpg - ERROR! "The connection was reset" (note.- added a space after http: to avoid link protection but this is not really promoting anything) I've been troubleshooting (3) for a couple hours and I'm unable to identify the culprit. I'm running NGINX 1.0.10 (latest stable) on Debian 6.0.2 32 bits. This NGINX instance runs another 40 or 50 sites with no problems.

    Read the article

  • Clicking or Knocking with Seagate HD

    - by Daniel A. White
    My laptop's main HD makes a clicking or knocking sound when Windows or the Bios tries to access it. I put it into a SATA dock and it sounds perfectly fine when spinning up, but after Windows tries to access it, it becomes a repetitive clicking or knocking sound. Does anyone know any tips that might help me access my data? I have most of it backed up, but I would still like to Ghost it before I send it off for repairs. I know my laptop is still under warranty.

    Read the article

  • Convert RAID-0 to RAID-1 on HP ML350G6 with P410i zero memory

    - by JLe
    I have an HP ML350 G6 with a P410i zero memory RAID controller. As far as I can understand that means I can't expand a current single drive "RAID-0" configuration to a RAID-1 using the HP Offline ACU without installing memory and BBWC. Is that correct? What makes me think about this is the fact that expanding RAID-0 to RAID-1 should be pretty similar to replacing a failed drive in an already existing RAID-1? So then why can't I expand without memory and BBWC? Is my best option otherwise to (i) use Ghost to capture the disk, create a new RAID-1 with the existing drive and a new one or (ii) buy memory+BBWC and do it online? Thanks

    Read the article

  • Home Server: cpu virtualisation, what to choose?

    - by Huygens
    I'm looking for virtualisation solutions for storage and OS for a home server. A sort of private cloud where I manage the storage space independently of the VM one. This question focus on VM (or compute instance) management and what would best suit my needs. (I have another question related to the storage management). My use cases are: A backup server: rsync and other services running. A personal cloud server: a kind of owned dropbox system, à la ownCloud. " users foreseen. A media server: streaming videos and displaying photos. Here my environement and wishes: Server: HP Proliant MicroServer with 8 GB RAM (AMD Turion dual core with AMD-V technology) OS types: only Linux (perhaps a *BSD VM in the future) Linux distributions do not matter, I'm familiar with RHEL, Fedora, Suse, Ubuntu, but any other recommandation will be fine 2-3 VMs foreseen: backup server, owncloud server and media server (optional). Those are only servers, so no graphical console needed (I don't need VirtualBox) By VM I mean a virtualised environment like KVM, Xen, etc. or a compute instance like with OpenStack storage should be "virtualised/cloudified" see my other question. VM should be able to be migrated to another server in the future if performance cannot be fullfilled anymore by the current server It does not matter if installation of such setup is complicated as long as management tools allow for easy maintenance I don't have Windows at home, so solution should be Linux friendly and would be nice to be web based. But native apps are OK too. System should be easy to enhance: by adding a new server to migate some of the VMs to it. So it's really a kind of private cloud on which I could run some Linux OS. I would prefer free (libre, as in a free speach) and open source tools. But it does not have to be free as in a free beer. So Xen, KVM, VitualBox or OpenStack? What would you recommend?

    Read the article

  • Implications of disabling the AMD Phenom's TLB patch?

    - by DMA57361
    I'm currently running a AMD Phenom X4 9600 processor (yeah, it's aging a bit, but other recent problems mean it's not getting upgraded in the immediate future), which happens to be one of the chips that suffer from the TLB errata. I recall that the first time I played with disabling the TLB patch (probably over a year ago, while playing a game that had a severe performance problem such that it was almost unplayable unless the patch was disabled) I had at least one BSOD, but I can't remeber them being particularly frequent. However, because it decreased instability, I stopped disabling the patch once I was done with the game. Now, after some recent hardware changes I was experiancing much worse performance than expected from the new hardware under some circumstances, and the TLB jumped to mind - after testing I found that disabling the patch would improve the performance to expected levels. I'm now wondering if it's worthwhile always having the patch disabled to avoid any potential slowdowns cropping up in the future, or if it is too dangerous. Everything I read states that the bug, when not patched, can causes a system lock-up in "rare circumstances". So, with the TLB patch disabled: How frequently should system lock-ups be expected? Do we know what the circumstances that trigger the lock-ups are? (Don't worry too much about being highly technical, but essentially I wonder if the chip more vunerable under heavy load, or heavy memory usage, etc?) Are there any secondary problems I should be aware of? (Don't include things that are charateristic to all lock-ups, please)

    Read the article

  • Cluster FIle System

    - by Ben
    We are looking for to choose a clustered file system for our in house appplication. Let me first highlight my requirement. we have a storage and 2 servers at present.We get the data files from remote servers to our server and on both servers we are running our application to access those data and make a final result as per our requirements. In future may be after 3-4 months, we can add another servers in current cluster pool to handle more data load from remote location data senders. So my requirement is that to integrate same storage partition on 2-3 servers , it might be 4-5 more servers in future, My application read data from storage partition and write back to storage partition. Is there any bottleneck / limitation from RHCS , GFS2 or anything.? We are new with RHCS + GFS and all. Can we have any other better approach or someway to deal with our requirement light way? what is the best OS version for this ? how's RHEL 6.4 64 bit ? please share some case study or some gudie reference as per past experiences with such environnmnets Regards, Ben

    Read the article

  • Backing up default windows installation with dd from linux running on another partition - is this fe

    - by Marek
    I am preparing to reinstall my system. I am thinking about creating a multi boot with a linux distro+Windows 7 to choose from when starting up. I would love to be able to skip all the hassle of reinstalling Windows and all programs when it starts becoming too slow in the future, thus I would like to mirror my fresh Windows system partition with some programs preinstalled. I am thinking about installing Ubuntu, making a partition for windows, installing windows with the basic environment (Visual Studio, Office, etc.) then booting into Linux and making an image of the windows partition with dd. I am not familiar with linux at all so I am a little afraid something may go wrong along the way. Is it possible to do it this way? Will I be able to partition my existing disk for multi boot easily after I install Ubuntu? Will I be able to recover the Windows partition easily using dd when I will need to re-create a fresh windows partition in the future? What other (better) approach can you recommend to achieve the goal of easy disk mirroring (for free)?

    Read the article

  • Anyone know a good mind mapper that works with a scheduler?

    - by GLycan
    TL;DR: Mind mapping tasks to be processed into a schedule based on task metadata. I have all sorts of ideas about what to invest resources (mainly time) in, but when I actually have time to do something I useually end up browsing reddit for not knowing what do to, and the frequancy with which I forget deadlines scares me. I'd love to bring order and structure into my mind, and always know what to do next. So, I want a mind mapping app, where I'd give each branch (types and subtypes of things I want to do) a importance score (if there were two branches, and one had 60 while the other 40, they would respectivily get 60% and 40% of the parent's importance, with the root being 100) and a how soon that branch should be revised/updated (an hobby I want to try out might be checked, say, once a week, while a school subject should be checked once a day) and give each leaf (something I want/need to do) how much time it takes, deadline (if any), and optionally an absolute importance, reoccurrence (guitar practice might repeat once a week), and prerequisites (reading something requires that book (although that could be brought somewhere), coding requires a box, jogging requires being outside) and maybe some other flags, like if it's enjoyable or not. It should either be packaged or working with a schedular app, to which I'd say, look, my day works this way (completely busy from 8 to 9:15, then 15 minutes of being inside with nothing, ..., two hours with box and possibility to go outside, etc), saying that such-and-such pattern is school and happens ever weekday except such-and-such days. The output should be of the form of a schedule, fit for printing or, when I finally get an android, mobile viewing, that schedules tasks with regards to availability of resources and importance (importance being derived from the leaf-task's parent branches), and the set of flags (all work and no play makes me a dull boy). One of these tasks should be reviewing anything that should be updated on that day, including future day layouts (e.g, if the time slots of future days have changed. This should be done every day.) Does anyone know some collection of preferably open-source (or free, or pirateable) tools, or better yet a single one, that accomplishes this task? I know python pretty well, and should be able to write any necessary glue.

    Read the article

  • Prevent registry changes by users

    - by graf_ignotiev
    Background: I run a small computer lab of 10 computers using Windows 7 x64 Enterprise. Our users are set up as limited users. For additional restrictions, I set up local group policy for non-administrators using the microsoft management console. Problem: Recently, I found out that some of these restrictions had been removed. Reviewing the settings MMC and in ntuser.pol showed that the settings should still be in place. However, the related registry settings were missing in ntuser.dat. I already have registry editing disabled in the GPO (though not in silent mode). Question: What is the best way to deal with this situation? Should I look into preventing registry setting changes? Should I set up registry auditing to found out how these keys are getting changed in the first place? Or should I give up the ghost and write some kind of logon script that enforces registry values if they've been change? Any other ideas?

    Read the article

< Previous Page | 47 48 49 50 51 52 53 54 55 56 57 58  | Next Page >