Search Results

Search found 10698 results on 428 pages for 'interactive session'.

Page 366/428 | < Previous Page | 362 363 364 365 366 367 368 369 370 371 372 373  | Next Page >

  • How do you optimize your Outlook Exchange + IMAP setup?

    - by Mike
    My company provides an Outlook/Exchange account we must use for mail/calendar. Like many companies, they unfortunately also provide a ridiculously small mail quota. I got tired of managing and backing up .pst files (since I'm always in my e-mail there is never a good time to back it up), so I started storing my archived mail "in the cloud", using an IMAP server I set up on my Linux box. This has a few drawbacks for me: IMAP (at least the implementation in Outlook) is *very slow*. Furthermore, if I move a large number of messages to the IMAP server, it blocks the entire Outlook client for hours sometimes, which is quite annoying. Can't use exchange over HTTP to do mail without launching a VPN session, because the client-side rules I have which organize my mail fail and disable the rule if the IMAP server can't be reached. If I reply to a message from my IMAP store, I have to specify a SMTP server willing to relay for me in order to send e-mail, unless I always remember to select my Exchange account while composing e-mail. ... but the main advantage of being very easy to back up, with a couple of cron jobs that essentially do an 'rsync'. Short of moving the IMAP server to my local host (which seem like might have the same file locking problems as using a .pst), my options seem limited for solving (1). I'd like to come up with a solution for (2) and (3) though. For problem (2) would it be possible to somehow tell Outlook that the IMAP server is "offline", and have it synchronize my changes during a periodic "send and receive"? If so, I wonder if it would block the Outlook client, like it does in problem (1), and if it would be compatible with the client-only rules I use to sort my mail into folders. I've looked all over the options menu and have not found a way to tell Outlook to not use a certain account for sending mail, which would solve (3). Is anyone else crazy enough to be doing something like this? Any ideas?

    Read the article

  • logfile deleted on Oracle database how to re-create it?

    - by Daniel
    for my database assignment we were looking into 'database corruption' and I was asked to delete the second redo log file which I have done with the command: rm log02a.rdo this was in the $HOME/ORADATA/u03 directory. Now I started up my database using startup pfile=$PFILE nomount then I mounted it using the command alter database mount; now when I try to open it alter database open; it gives me the error: ORA-03113: end-of-file on communication channel Process ID: 22125 Session ID: 25 Serial number: 1 I am assuming this is because the second redo log file is missing. There is still log01a.rdo, but not the one I have deleted. How can I go about recovering this now so that I can open my database again? I have looked into the database created scripts, and it specified the log02a.rdo file to be size 10M and part of group 2. If I do select group#, member from v$logfile; I get: 1 /oradata/student_db/user06/ORADATA/u03/log01a.rdo 2 /oradata/student_db/user06/ORADATA/u03/log02a.rdo 3 /oradata/student_db/user06/ORADATA/u03/log03a.rdo 4 /oradata/student_db/user06/ORADATA/u03/log04a.rdo So it is part of group 2. If I try to add the log02a.rdo file again "already part of the database". If I drop group 2 and then add it again with these commands: ALTER DATABASE ADD LOGFILE GROUP 2 ('$HOME/ORADATA/u03/log02a.rdo') SIZE 10M; Nothing. Supposedly alters the database, but it still won't start up. Any ideas what I can do to re-create this and be able to open my database again?

    Read the article

  • 'The rpc server is unavailable' or 'access is denied' error when using Remote desktop Services Manager on Windows 7 (but mstsc.exe works!)

    - by tbone
    I am trying to connect to a Windows XP workstation from a Windows 7 Ultimate workstation using Remote Desktop Services Manager. I am able to do a Remote Desktop (mstsc.exe) session from the Win7 machine to the WinXP machine with no problem at all. When running the Remote Desktops Admin (tsmmc.msc) too on a Windows XP box, I can also connect with no problem. However, when I use the new Remote Desktop Services Manager on Windows 7 and try to connect, I get the error: "The rpc server is unavailable" What could cause this? Has there been some fundamental change in Remote Desktop Services Manager, does it connect in a different way somehow? Update #1 Turned off firewall on the Windows XP box and the "The rpc server is unavailable" error went away; so RDSM seems to be using an entirely new port/connection/service compared to mstsc.exe or the old Remote Desktops Admin tool. Now... after disabling the firewall, I get a new error: Access is Denied. After doing some googling, I found some articles discussing this; basically, the error is very misleading - the actual problem is, if either side of the connection has dual monitors, and they are not both Win7 Ultimate, then you cannot connect using Remote Desktop Services Manager...the reason is, by default it uses the /multimon switch, and this switch requires a certain level of Windows license - and, there seems to be no way of changing this default (if anyone knows of a way to change this default, please post an answer or comment!). Nice going Microsoft. http://social.technet.microsoft.com/Forums/en-US/windowsserver2008r2rds/thread/4d06278f-e0f4-4f8e-a8e1-3697ee967ef4 http://www.experts-exchange.com/OS/Microsoft_Operating_Systems/Windows/Windows_7/Q_26225743.html

    Read the article

  • Clicking far away in vim in tmux in urxvt

    - by paps
    I use vim inside tmux inside urxvt, and the mouse works perfectly well for clicking and selecting text, except when I want to click too far to the right. It seems to be related to the distance in number of columns from the left. When I go beyond column ~200 (not sure about the exact number), clicking simply does nothing. Note that it's not related to a vim window: with two vim windows taking ~150 columns each, clicking will not work after the ~50th column in the second window. It's related to the whole vim session. Also note that clicking far away in a big tmux pane (200 columns) works perfectly. In my .tmux.conf I have this line: set -g default-terminal "screen-256color" and in my .vimrc I have this: if &term =~ "^screen" autocmd VimEnter * silent !echo -ne "\033Ptmux;\033\033]12;7\007\033\\" let &t_SI = "\<Esc>Ptmux;\<Esc>\<Esc>]12;5\x7\<Esc>\\" let &t_EI = "\<Esc>Ptmux;\<Esc>\<Esc>]12;7\x7\<Esc>\\" autocmd VimLeave * silent !echo -ne "\033Ptmux;\033\033]12;14\007\033\\" end It changes the cursor's color depending on the editing mode of vim, and it works, meaning that tmux really sets $TERM to "screen-256color" — but I don't know if this has any relevance with my mouse problem. I'm running Ubuntu 12.04, vim 7.3, tmux 1.6 and rxvt-unicode 9.14. Does anybody have an idea about what is causing this problem? Thanks.

    Read the article

  • How is network mounted software executed?

    - by CptSupermrkt
    I would like to understand how network mounted software works. For example, at my place of work, we have a software server. Each client machine (hundreds of them) automatically mounts directories from the software server on boot. For example, a program like Matlab is installed just once on the software server, but each client machine can start up an instance of Matlab. What is going on under the hood? Let's say I run /opt/bin/matlab and /opt/ is mounted from the software server, what happens when I press Enter to execute matlab on a client machine? The process is on the client machine, and I've already narrowed down that there isn't any implicit or hidden file transfer (i.e. copying matlab to my machine temporarily for that session) by running matlab on a computer with nearly zero disk space (i.e. not enough room to transfer). Since Matlab was installed on the server, how is my client computer executing it? What mechanism is controlling this? What is happening behind the scenes?

    Read the article

  • Dead Linux server - need help and options

    - by Choi S.
    All, I have a Dell PE 1950 w/ 2 SATA drives in a software RAID1. OS is CentOS 5.5 (2.6.18.x). Starting this afternoon we received HW errors (something on the bus is bad, E171F) and the machine became unresponsive. We hard booted and it came back up for about 5 hours but then it happened again. I'm trying to figure out our options. Unfortunately we do not have similar hardware but I have a small desktop that I can use. I was contemplating putting one of the drives into the desktop and then starting it up. My goal was to then P2V it using Vmware converter but apparently the free v5.x doesn't support hot cloning/converting on a RAID volume, only the Enterprise 4.x version of Converter does. My questions are: 1.) Is putting a single drive out of a RAID1 pair into another piece of HW is safe? Based on my research and understanding it appears to be but would like confirmation. 2.) Is there any work around to the Vmware Converter not supporting RAID volumes during a hot clone/convert session? 3.) Are there other options I'm overlooking? Thanks in advance for reading and responding. --Choi S.

    Read the article

  • printer assignments for windows xp workstations within an active directory environment

    - by another_netadmin
    I'm using the following script to handle removing any old networked printers from machines and then assigning the propper ones and making one of them the default. This script is assigned to the OU the workstations reside in and uses group policy loopback so all users that login will get the appropriate printers mapped for them. I tried to use the new Printer Manager as part of W2K3 R2 but when assigning the default this way I get an error that the printer doesn't exist so I'm back to using the script. One flaw that I'm noticing is that it won't remove any printers that happen to be mapped from an RDP session (we don't see this everywhere but there are a few locations). Is there any way to enumerate all RDP printers and remove them similar to how I'm enumerating and removing networked printers? ' ' Printers.vbs - Windows Logon Script. ' RemovePrinters AddPrinters Sub RemovePrinters() On Error Resume Next Dim strPrinter Set objNetwork = WScript.CreateObject("WScript.Network") Set colPrinters = objNetwork.EnumPrinterConnections For i = 0 to colPrinters.Count -1 Step 2 strPrinter=CStr(colPrinters.Item(i+1)) If Not InStr(strPrinter,"\\") = 0 Then objNetwork.RemovePrinterConnection strPrinter, True, True End If Next End Sub Sub AddPrinters() On Error GoTo 0 Set objNetwork = CreateObject("WScript.Network") objNetwork.AddWindowsPrinterConnection "\\printers1\JH120-DELL5310" objNetwork.SetDefaultPrinter "\\printers1\Jh120-DELL5310" End Sub

    Read the article

  • tomcat vs FULL J2EE Solutions

    - by jrhickey
    We are getting ready to make a major revision in our Web Application architecture which currently is running on JBoss 4.2. At first we were looking at moving from 4.2 to JBoss 6 but after some research tomcat may be a better solution for us. My first question is their anything that JBoss can do that tomcat cannot do assuming you are using the correct plugins. We do not really use EJB's in our solution and it would appear there are simple plugins for web services, JMX and other features. Tomcat appears to have much better support, faster upgrade cycles and many, many books. Since there is less to the system it also seems much easier to support from an admin point of view. What am I missing? The main features we want to enable are better clustering support and session replication / persistence. We will consider other application servers as well such as Glassfish / Geronimo. quote form a web article: Apache Tomcat is the world’s most widely used web application server, with over one million downloads per month and over 70% penetration in the enterprise datacenter. Tomcat is used to power everything from simple one server sites to large enterprise networks.

    Read the article

  • rdp allow client reconnect without password prompt after several hours

    - by Tom
    Let me describe the setup first: client PC with several rdp sessions to local servers, all opened from saved rdp sessions with stored passwords, using the standard windows rdp client. several windows servers on the LAN, with varying server OS: windows server 2003, 2008, and even 2012 now. When I log onto my PC I open up rdp sessions to all those servers, and keep them open all the time for various reasons. Overnight the client PC is put into sleep or hibernate mode, thereby braking the rdp connections. On the next day when I wake the client PC and login again, the rdp sessions automatically try to reconnect to the servers, and this leads to the question: starting with server 2008 something apparently changed in the rdp server config, as all servers with 2008, 2008r2 and 2012 will prompt for the password in the rdp session, whereas the 2003 server rdp connections will re-establish without the password prompt. Apparently there is a timeout setting on 2008+ that, when exceeded, requires a reauthentication. Is there any way to setup the 2008+ servers to behave like 2003 did? I'd like the rdp sessions to reconnect without a password prompt even after a several hour disconnect.

    Read the article

  • Office 2010 Trust Center settings: How to enable data connections in the "old" way?

    - by GSerg
    We're planning an upgrade Office 2003 - 2010 and have identified a big problem. In Office 2003, if the workbook you're opening contains a query table that fetches data from a data source automatically (upon file open or in certain intervals), then a security dialog pops up - whether you want to allow that. If you say Yes, the queries will refresh automatically when they need to. If you say No, the queries will not refresh automatically, neither on file open nor on time intervals, but you will be able to refresh any of them manually at any time by right-clicking and selecting Refresh. There is also a registry parameter to say, Don't display that dialog, just allow the queries. This is exactly what we want. On users' computers we have the registry parameter applied, so the users never see any dialogs. On developers' computers the parameter is not applied, so every time a file is opened the developer decides whether to allow the auto-refreshing for the current session. Usually the answer is No, because for developing, it is essential to not have quieres refresh when they want to, but instead, refresh them when the developer wants. The problem is that in Office 2010 which we are testing we can't find a way to achieve this functionality: The allow/disallow messages are now grouped into one yellow button, that either allows everything or disallows everything (including, say, macros, if macro security is set to "Disable, but ask"). If you don't click the yellow Allow button, the queries are disabled completely, not just for automatic execution. You cannot right-click and refresh a particular query -- doing that would summon a security dialog prompting for enabling queries, and if you say Yes, all queries in the document will be enabled for auto-execution and will start executing immediately. This sort of ruins our development environment. Is there a way to get the trust thingies in Office 2010 to work in the same way as before? Is there a yet another registry parameter to say, Prompt for auto-refresh, but allow manual refresh even when auto-refresh is disabled?

    Read the article

  • REMOTE_USER not getting set?

    - by landed
    I am trying to setup LDAP Authentication in Joomla using a plugin called JMapMyLDAP (in fact 4 plugins each doing a different job). I need to pull a part of a string out of the server variable REMOTE_USER and this should be visible (we see here http://timplummer.com.au/4-how-to-integrate-joomla-3-with-active-directory-using-ldap.html) in phpinfo(); The issue is that REMOTE_USER is not set or at least not appearing. A few things to note (if you don't mind) here- conceptually I am not really understanding authentication as a whole subject it appears to be vast despite my years working in websites. Yes I used asp and built php pages to check a user is who they say they are with a token(/session?) that was given to just them and then they are identified when a stateless request is made to the server. Thats my level of understanding. This sounds different to the basic authentication in apache where a password sits in a file and a username and the user needs to login to a basic form to get access to the folder/docs this is via an .htaccess file. Ok so with the LDAP to work I need to get REMOTE_USER this sounds very reasonable as how else do we know is making the request. Thank you.

    Read the article

  • What are the possible problems, when wget returns code 500 but same request works in normal browsers?

    - by markus
    What should I be looking for, when wget returns 500 but the same URL works fine in my web browser? I don't see any access_log entries that seem to be related to the error. DEBUG output created by Wget 1.14 on linux-gnu. <SSL negotiation info stripped out> ---request begin--- GET /survey/de/tools/clear-caches/password/<some-token> HTTP/1.1 User-Agent: Wget/1.14 (linux-gnu) Accept: */* Host: testing.thesurveylab.net Connection: Keep-Alive ---request end--- HTTP request sent, awaiting response... ---response begin--- HTTP/1.0 500 Internal Server Error Date: Wed, 12 Dec 2012 14:53:07 GMT Server: Apache/2.2.3 (CentOS) Set-Cookie: blueprint2-staging=8jnbmkqapl30hjkgo0u6956pd1; path=/ Expires: Thu, 19 Nov 1981 08:52:00 GMT Cache-Control: no-store, no-cache, must-revalidate, post-check=0, pre-check=0 Pragma: no-cache Strict-Transport-Security: max-age=8640000;includeSubdomains X-UA-Compatible: IE=Edge,chrome=1 Content-Length: 5 Connection: close Content-Type: text/html; charset=UTF-8 ---response end--- 500 Internal Server Error Stored cookie testing.thesurveylab.net -1 (ANY) / <session> <insecure> [expiry none] blueprint2-staging 8jnbmkqapl30hjkgo0u6956pd1 Closed 3/SSL 0x0000000001f33430 2012-12-12 15:53:07 ERROR 500: Internal Server Error.

    Read the article

  • troubleshooting really slow login on a (linux) machine

    - by Peeter Joot
    Within the last couple of weeks, any attempt to login to a specific linux server has gotten really slow. Once I've logged in, things appear to run without significant delay, but some other login like activities (like starting a new screen session) are slow. The machine's been rebooted a couple of times recently and that hasn't helped. , and it doesn't appear to be $PATH search (where $PATH can sometimes include bad NFS mounts), which I've seen historically in our environment. I've also tried completely removing my .profile/.bash*/... type of init files to rule out anything bad there. I also see slow login for at least one other userid on the system. One thing I've noticed is the following message when trying to exit from a screen terminal: Utmp slot not found -> not removed and am wondering if this is related (having a vague recollection that Utmp has something to do with login). Any idea what that message means, or how to fix it, and if it would be related? Failing that, what sort of problem determination tools are available to investigate what is slowing down this login process?

    Read the article

  • Debugging nginx URL rewrite: How do I figure out where the problem is?

    - by pjmorse
    I have a specific URL pattern on a site which needs to be redirected to the HTTPS version. This is a Django site; Nginx checks each URL in memcached, and if it doesn't find a cached version it proxies the request to Apache/mod_python for Django to render the page. The relevant configuration block is rewrite ^/certificate https://mysite.com/certificate ; rewrite ^/([a-zA-Z]{2})/certificate https://mysite.com/certificate ; ...and it doesn't appear to be working at all. Nginx is: $ nginx -V nginx version: nginx/0.7.65 built by gcc 4.2.4 (Ubuntu 4.2.4-1ubuntu4) TLS SNI support disabled configure arguments: --prefix=/usr/local/nginx --pid-path=/var/run/nginx.pid --with-http_gzip_static_module --with-http_ssl_module How can I figure out if the problem is my patterns not matching, or a more obscure configuration problem? (The site is localized to three languages, and the localization is in the URL string, e.g. /US/news/, /DE/about, etc. It tracks localization in the session as well, defaulting to US, so if you just requested /news Django will rewrite to /US/news unless the user has a cookie indicating they're using a different localization. Django handles this, though, not Nginx.)

    Read the article

  • Extend Linux Desktop to another X Windows Display

    - by unknown (google)
    Hello, I am a long time Linux user of the Xinerama and other technologies for extending a desktop to multiple monitors. However when I travel with my laptop I miss the multi-monitor support I enjoy at home. Recently I acquired a second laptop for a low price. Both laptops are running Fedora (versions 10 and 11 respectively). I use Gnome as my primary desktop environment. I know about synergy. I use synergy all the time to control the screen of other Windows / Linux systems I use. I would like to know, can I sit both my primary and secondary laptops together and achieve a Xinerama-like extended desktop environment? Ideally I would like to start a GNOME session on my primary laptop. And then start a X-Windows Desktop on my secondary laptop and extend my primary laptop's desktop onto it. I would like to be able to move Windows from the primary desktop to the secondary laptop desktop. Would I need to use synergy to do this with some other bit of X-Windows technology? Or is there X-Windows technology that will do all this for me? I am familiar with X Windows ability to display applications remotely. I am also familiar with Nomachine's NoX.

    Read the article

  • How does one debug Windows network share authentication?

    - by ajs410
    I have machine0 with 32-bit Vista, logged in as a domain user, running a VMWare image of 32-bit Vista, logged in as a local user, with the VM set to bridge the network. From an administrator account (called admin) within the VM, I try to access the hidden C$ share on machine0 (i.e. start - run - "\\machine0\C$\"). I get no prompts for credentials. Worse, machine0 has an admin account (different password), and machine0\admin gets locked out when VM\admin tries to access the network share. I get a message several seconds later, which feels like a cached credential failure leading to the lockout. I have checked several places for cached credentials; net use, Stored Usernames and Passwords, mapped shares. I rebooted (both machine0 and VM) to make sure the session was clear of any cached credentials. I can force net use to use my domain credentials when accessing machine0, and then I can see the share. I can also see shares that do not require credentials. I decided to try another machine on the network (machine1), 64-bit Vista, local user. This machine has no lockout policy, and after several seconds (feels like failed cached credentials again) it prompts me for credentials. After I enter them, it re-prompts me, saying "logon unsuccessful" (tried my domain credentials, and also machine1\admin's). Which is bogus, because I proceed to log on with remote desktop using the machine1\admin credentials. I have tried this on another machine (machine2, 64-bit Vista), running a copy of the same 32-bit VM, and I don't remember having this problem. machine0 has a fingerprint reader...could that try storing passwords and interfere? Are there any places I'm missing where there could be cached credentials? Is there a way to see what credentials are flying around when I try to connect?

    Read the article

  • Repair Windows 2008 boot

    - by aksamit
    I have been caught in a loop where I can't boot up Windows Server 2008. When the OS starts to boot I am presented two options: Launch Startup Repair (recommended) Start Windows Normally No matter what option is chosen I end up with a black screen stating "Windows is loading files..." which takes approx 15 minutes until "System Recovery" is started. I try to run the "Startup Repair" but it reports the following error: Problem Event Name: StartupRepairOffline Problem Signature 01: 0.0.0.0 Problem Signature 02: 0.0.0.0 Problem Signature 03: unknown OS Version: 6.1.7600.2.0.0.256.1 Locale ID: 1033 Actually no matter what other options I would try in the System Recovery they would also report some kind of error. Some other logging I have been able to extract: Diagnosis and repair details: Number of repair attempts: 1 Session details System Disk = Windows directory = AutoChk Run = 0 Number of root causes = 1 Test Performed: Name: System disk test Result: Completed successfully. Error code = 0x0 Root cause found: A hard disk could not be found. If a hard disk is installed, it is not responding. So it is actually stating that Harddisk could not be found... which to me seems to be a bit contradictious since the System Recovery just loaded all files from the hard disk it now states it cannot find. Any ideas on what other options I have?

    Read the article

  • Logging Remote Server Access via Remote Desktop

    - by Nate Bross
    The objective here is to start a simple .NET application I've written which captures some environment variables (time, username, computername, etc) upon login. This .NET application subscribes to the Windows "User logout" event. Upon launch, the application captures the above variables, and creates a record in my database, upon logout (which I'm capturing) I update another field in the same record, with the logout time. The above is working exactly as I would like, when I launch the binary, it makes its initial log entry, then waits for the logout event and updates the same record. Restrictions, the .NET binary should be able to live on a share point (\server\share\myapp\v1) so I can update the application to (\server\share\myapp\v2) and simply update the GPO/Logon script. My initial thought was to use the \domaincontroller\sysvol\ directory to store the binary and then update all user accounts to include a call to my application. Can you see any flaws in this approach? My question is this: First, is there anything wrong with my idea above? Second, if so, what is the best way (through group policy or otherwise) to ensure this application launches whenever a session is started on a server?

    Read the article

  • How can I rapidly switch hosts?

    - by EAMann
    I'm in the process of migrating a forum setup from one version of the software on one machine (older shared Windows host) to a new VPS (Windows Server 2008). To install the software, I used my hosts file to temporarily point the domain at the new IP address. To see the old site, I obviously re-edit the hosts file to remove the reference. But this leaves me constantly adding/removing a # from my hosts file just so I can switch back and forth between the two servers. Is there a way to do this more rapidly? I've found a handful of toggling batch scripts, but all they do is automate the addition/removal of the # character ... so there's still a noticeable lag where I have to repeatedly hit F5 to force my system to detect the new settings. Ideally, I could view both servers at the same time on the same machine. Maybe one through a regular browser session and one through some kind of a proxy. Unfortunately, I don't have the first idea how to set that up. Ideas?

    Read the article

  • Screen randomly goes blue/black/white

    - by FubsyGamer
    Problem Randomly, while using my computer, the monitor goes dark grey/almost black, or it goes white with faint grey vertical lines, or it goes blue with black vertical lines. It's as if the computer powers off. People tell me I sign out of Skype, Spotify stops playing when it happens, etc. When I look at the tower, it doesn't seem like it's off at all. Nothing changes, fans are spinning, lights are on, etc. If you were only looking at the tower, you'd never know there was a problem The only way I can get it to come back up is to push and hold the power button and turn it off, then back on This never happens while I'm playing video games. I've done 5-6 hour sessions of League of Legends, and it doesn't do anything When I'm just browsing the web, reading email, checking Reddit, etc, it happens all the time. It can happen multiple times in a session, it usually takes only about 5 minutes from the time I start browsing to when the computer crashes This started happening after I moved to a new apartment (this has to be relevant somehow, it was not happening where I lived before) There is nothing in the crash logs or event logs System Specs i5 2500k CPU AMD Radeon 6800 GPU Gigabyte z68a-d3h-b3 motherboard WD VelociRaptor 1 TB HDD Screenshots Device manager About screen Things I have tried I was getting a WMI Error in my event logs, but I fixed it using Microsoft's fix, KB 2545227 I was using Windows 8. I wiped the HDD and downgraded to Windows 7 64 bit I took out the video card and used a can of air to totally clean out the video card, all fans, and the inside of the computer in general. I made sure all of the video card pins were fine, then reconnected it I tried to update my motherboard BIOS, but anything I downloaded from Gigabyte was only for 32 bit machines, not 64. I don't even know how to tell what my motherboard BIOS is at right now I am using a power strip, and anything else connected to it works just fine If I re-seat the monitor cable while this is happening, nothing changes Please, help me. I've been battling this for several weeks now, and it's so frustrating it makes me not even want to use the computer.

    Read the article

  • Is my "Generic" USB Flash Drive broken?

    - by Jesse J.
    So here is the situation. I find myself technological knowledgeable about many things ( I love to code, whether it's websites, C#, C++ or so on). However: My 2 toddlers (my wife actually) bought me a "Generic" 128 GB USB Storage Device (Usb Flash Drive) for Father's Day. I thought awesome at first..... WRONG! Nothing but problems with it. 3-4mb/s MAX transfer speed. I can bear with it. BUT! When I went to reformat my computer I transferred my save files from my games over to the stick and then the USB Stick managed to become corrupted. Not just a simple format would work either. It's screwed. I tried to use (Manually changed usb drive letter troubleshooting it to X) "chmod X: /X /F /R" with administrator rights, I did this after a long session to make it work with no errors (had to delete the log) and I finally recovered the files, however when I go to use it (transfer to or from) it transfers a couple kb to the stick or from it and then freezes, It says (Windows 7): Name: From: Folder (X:\File\Location) To: Folder C:\Users\Username\Desktop) Items Remaining: 0 (0 bytes) Speed: 0 bytes/second It does this forever... and ever... and ever... It transfered 3 files atleast, and then stopped. This is a new USB Stick bought from a "High" reputation company on eBay. Is the USB Stick screwed?

    Read the article

  • Scaling a node.js application, nginx as a base server, but varnish or redis for caching?

    - by AntelopeSalad
    I'm not close to being well versed in using nginx or varnish but this is my setup at the moment. I have a node.js server running which is serving either json, html templates, or socket.io events. Then I have nginx running in front of node which is serving all static content (css, js, etc.). At this point I would like to cache both static content and dynamic content to memory. It's to my understanding that varnish can cache static content quite well and it wouldn't require touching my application code. I also think it's capable of caching dynamic content too but there cannot be any cookie headers? I do use redis at the moment for holding session data and planned to use it for other things in the future like keeping track of non-crucial but fun stats. I just have no idea how I should handle caching everything on the site. I think it comes down to these options but there might be more: Throw varnish in front of nginx and let varnish cache static pages, no app code changes. Redis would cache dynamic db calls which would require modifying my app code. Ignore using varnish completely and let redis handle caching everything, then use one of the nginx-redis modules. I'm not sure if this would require a lot of app code changes (for the static files). I'm not having any luck finding benchmarks that compare nginx+varnish vs nginx+redis and I'm too inexperienced to bench it myself (high chances of my configs being awful). I'm basically looking for the solution that would be the most efficient in terms of req/sec and scalable in the future (throw new hardware at the problem + maybe adjust some values in a config = new servers up and running semi-painlessly).

    Read the article

  • Logging Remote Desktop to Servers via Logon Script or GPO or What?

    - by Nate Bross
    The objective here is to start a simple .NET application I've written which captures some environment variables (time, username, computername, etc) upon login. This .NET application subscribes to the Windows "User logout" event. Upon launch, the application captures the above variables, and creates a record in my database, upon logout (which I'm capturing) I update another field in the same record, with the logout time. The above is working exactly as I would like, when I launch the binary, it makes its initial log entry, then waits for the logout event and updates the same record. Restrictions, the .NET binary should be able to live on a share point (\server\share\myapp\v1) so I can update the application to (\server\share\myapp\v2) and simply update the GPO/Logon script. My initial thought was to use the \domaincontroller\sysvol\ directory to store the binary and then update all user accounts to include a call to my application. Can you see any flaws in this approach? My question is this: First, is there anything wrong with my idea above? Second, if so, what is the best way (through group policy or otherwise) to ensure this application launches whenever a session is started on a server?

    Read the article

  • Remote Desktop leaves host unresponsive

    - by Jeff Dalley
    I have my desktop PC at home set up to accept remote connections, and I often connect to it from work on my laptop via mstsc.exe. However, every time I remote to it, I find when I go home that despite the monitor being on - it's not receiving an image and it looks as though the computer is hibernating/asleep. I basically have to restart it whenever I get home and I know there's an answer for why its doing this. More details: When exiting the remote session, I have tried both logging off the account, and closing the RDP window without logging off; both give the same result. When I get home to the desktop I of course try moving the mouse, ctrl+alt+del to see if its responsive to restart, multiple key-press to see if I can get any audio out of it; It seems pretty obvious its sleeping/hibernating in some way: Nothing happens in any of these cases and a physical restart is necessary. Both desktop and laptop are running Windows 7 Ultimate. I'm thinking it really is sleeping/hibernating it, and I'm not sure why because left alone my desktop's power options are set to never turn off the HDD or change its state - I leave it on 24/7. This could be a stupid error on my part but I just can't see it! Thanks.

    Read the article

  • Network problems that might be related to NAT

    - by nenne
    Hello, I have an odd setup where there is a router(Router 2) routing between network network 1 and network 2. One router(Router 1) with nat for internet access that routes between internet and network 1. There are people in both of these networks. All the clients in network 1 can access the internet, the clients in network 2 can access the clients in network 1 and can also access the router 1. Router 1 can also access clients in network 2. However, the clients in network 2 cannot reach the internet. I cannot think about anything in the routing tables that would hinder this, since Router 1 can reach the clients in network 2 and vice versa. Can it be that nat starts the session between router 2 and the internet site/machine instead of the client and the internet machine? Does anyone have any ideas? I have very little control over router 2(its basicly an ISP vpn net service) but full access to router 1. Its an ubuntu 10.04 with iptables for nat/firewall setup.

    Read the article

< Previous Page | 362 363 364 365 366 367 368 369 370 371 372 373  | Next Page >