Search Results

Search found 3106 results on 125 pages for 'talking shoes'.

Page 93/125 | < Previous Page | 89 90 91 92 93 94 95 96 97 98 99 100  | Next Page >

• Half of installed RAM is hardware reserved

  - by user968270

  After a rather arduous and convoluted series of problems that left me without a desktop for ~80 days, I've finally got the thing up and running, having replaced the power supply, motherboard, graphics card and CPU. Now, however, I'm experiencing the 'hardware reserved RAM' issue. Perhaps this is the exhaustion talking, but looking at the question that tends to get pointed to when this kind of topic gets locked as a duplicate hasn't helped. I have 16 GB of RAM installed in an MSi 970A-G46, which is spec'd for up to 32 GB of RAM. The BIOS recognizes that I have 16 GB installed, and the resource monitor also shows the whole 16 GB, only it shows 8 GB as hardware reserved. I've seen suggestions that it's an OS issue, but the particular installation of Windows 7 (64-bit) which I'm running on my boot drive is the same as the one that could actually access the 16 GB in my previous motherboard (MSi 870A-G54). I've updated my BIOS using the MSi Live Update tool and restarted the machine with no effect, and I cannot seem to locate any 'Memory Remapping' option as I've seen mentioned. I've physically swapped the RAM between the slots to no effect. I've unchecked the Maximum Memory box in the msconfig Boot tab's advanced options, also to no effect. These are my system's basic specifications OS: Windows 7 Home Premium (64-Bit) Motherboard: MSi 970A-G46 CPU: AMD FX-8150 Graphics Card: XFX Radeon HD 6870 Boot Drive: OCZ Agility 3 Storage Drive: Samsung Spinpoint F3 ST1000DM005/HD103SJ 1TB PSU: Thermaltake TR-2 TR600 600W ATX12V v2.3

  Read the article

• How to turn off Windows Azure's "This copy of Windows is not genuine" message?

  - by Sid

  Is there any setting/configuration item to avoid Windows Azure from printing that error on the screen or detecting it? I've put a screenshot below that shows the message when you RDP into the web role. My web role runs on Windows Azure Guest OS 1.17 (variant of Windows Server 2008 SP2) Background: I was explaining our architecture to some outside engineers (NDA'd and all) and had to demystify the webrole as they were unfamiliar with Azure. I RDP'd into the VMs running the Web Role when one of their engineers gasped "are you guys running pirated copies of Windows in the cloud?" I also noticed that within the RDP screen, the Azure machines had "This copy of Windows is not genuine" on the bottom left corner. Now obviously, Microsoft is running their own OS in their own datacenter with no influence from me. So no 'piracy' here, despite that obvious warning. However, they seemed so distracted by this ("how can it be? really? hmmm?") that we wasted more time talking about it than the actual matter at work. Like I said, they have little exposure to Azure but have value add elsewhere. I want to get rid of this so I don't have to explain this in the future. PS Microsoft: If you're going to modify Windows Server <XYZ> into Windows Azure <A.B> , you should also modify the code that verifies product integrity.

  Read the article

• Ubuntu 12.04 Very slow especially with Android Studio

  - by Nada

  I have an old laptop with the following specification: Memory: 485 MiB, Processor: Genuine intel CPU T2300 @ 1.66 GHz ×2, OS Type: 32 bit, Disk: 78.1 GB, I installed on it Ubuntu 12.04 LTS and I noticed that the overall system is very slow in responding. I tried to search about that in the internet and I found some articles talking about how to make Ubuntu 12.04 LTS run fast I applied all what they said including download LXDE desktop environment and then nothing different in the system response time. Then I need to develop some android applications so, I download Android Studio (Beta) 0.8.6. The problem became worse than before whenever I tried to open the Android Studio the screen is frozen for some minutes then it took time to download the projects and initialize the work space also, when I tried to move the cursor he is move very slowly. When I tried to run my first application on the AVD it took three hours and still not run yet. I delete the Android Studio and install it again several times, I was trying to solve the problem but still nothing change. Please if you have any suggestions that may help me make my laptop and Android Studio work faster I will appreciate it for you. Thank you in advance.

  Read the article

• Securing NTP: which method to use?

  - by Harry

  Can someone good at NTP configuration please share which method is the best/easiest to implement a secure, tamper-proof version of NTP? Here are some difficulties... I don't have the luxury of having my own stratum 0 time source, so must rely on external time servers. Should I read up on the AutoKey method or should I try to go the MD5 route? Based on what I know about symmetric cryptography, it seems that the MD5 method relies on a pre-agreed set of keys (symmetric cryptography) between the client and the server, and, so, is prone to man-in-the-middle attack. AutoKey, on the other hand, does not appear to work behind a NAT or a masquerading host. Is this still true, by the way? (This reference link is dated 2004, so I'm not sure what is the state of art today.) 4.1 Are public AutoKey-talking time servers available? I browsed through the NTP book by David Mills. The book looks excellent in a way (coming from the NTP creator after all), but the information therein is also overwhelming. I just need to first configure a secure version of NTP and then may be later worry about its architectural and engineering underpinnings. Can someone please wade me through these drowning NTP waters? Don't necessarily need a working config from you, just info on which NTP mode/config to try and may be also a public time server that supports that mode/config. Many thanks, /HS

  Read the article

• Monitoring instantaneous network throughput at one second intervals?

  - by Shaddi

  For a testing setup I have, I need to monitor the throughput through a "router"* at regular intervals of around 5 seconds or less (sub-second intervals would be very nice, but not required). Ideally, I would be able to generate a file which contained both the number of bytes and packets seen during each interval. I will eventually be generating a time-series of throughput from this data. On a previous setup using an older version of FreeBSD, there was a tool called "bpfmon" which gave me this information. However, I need to do this under a modern version of Linux (namely, Ubuntu 11.04). I have looked at both iptraf and iftop, but these do not appear to provide the resolution I need, nor do they seem to easily allow scraping the data I need. I understand iptables statistics may be able to give me what I'm after, but the examples I've seen of this seem to rely on repeatedly reading and resetting traffic counters, which seems like it could give inaccurate as read/reset is not an atomic operation. I already capture a tcpdump trace of the traffic I'm interested in on the link I want to monitor, so I am open to approaches which simply parse that. I feel like this must be a common problem though, so I am hoping there will be a standard "best practice" tool for accomplishing this. *I say "router" in quotes because I am really talking about a machine with two bridged NICs through which all the traffic I'm interested in passes.

  Read the article

• Is it possible to record a screen-video from a VNC server?

  - by nikie

  I have a computer that's running VNC server. I would like to record a video of what's going on on this computer, if possible without installing additional software on that computer. Is there a program that can connect to the VNC server port and instead of displaying the screen save it to an (e.g. AVI) video file? Background: One of our customers sometimes has problems with the software he bought from us when he's performing a complex procedure. To help him, we offered that someone (a service technician or programmer) watches what he's doing during that procedure to find out if he's doing something wrong or if there's a bug in the software. Currently, this is done live via VNC. That has a few disadvantages: The service technician has to be in the office at the time. As the customers are in different time zones, that can be in the middle of the night. If the service technician forgets something or doesn't notice something, it's lost. There's no way to see what happened again. Only a single computer can be watched by one service technician at a time. I know I could install normal screen-grab software on the computer, but we're talking about an embedded system with limited RAM, CPU, HDD space, so installing something new is not an easy decision. And VNC is already there. I could of course open a VNC client on some office PC and capture that PC's screen, but I can only record one remote computer that way. I often have to watch up to 8 screens in parallel. (And I don't think that screen-grabbing VNC would improve image quality, either.)

  Read the article

• How can I disable 'natural breaks' in Workrave?

  - by Pixelastic

  I've just discovered Workrave, and was trying to use it along the Pomodoro technique (5mn break every 25mn). But the concept of 'natural breaks' of Workrave seems to interfere with what I'm trying to achieve. Workrave tries to guess that I'm doing a natural break if I stop using my mouse and keyboard for longer than 5s. It then stops the work timer, and start counting time as if I was doing my break. Here is a typical example : I've configured a 5mn rest break every 25mn. I start working. 10mn later, I receive a phone call, or start talking with a colleague, or any work-related action that do not need either keyboard nor mouse. Workrave then stops counting my time as work time, and starts its rest timer. If my phone call is shorter than 5mn, then Workrave will resume its timer where it stopped it. Meaning that my time on the phone is not counted as work time, and so my break time is pushed a few minutes later than it should be. Even worse, if my phone call is longer than 5mn, then Workrave count it as a complete rest break, and when I'll resume working, it will restart its timer completly. I'm looking for either a way to disable the natural breaks, or increase the 'inactivity time' from 5s to maybe ~1mn. Or maybe an other angle to look at the natural breaks that might work with the Pomodoro technique (forced 5mn breaks every 25mn). I'm using Ubuntu 11.10.

  Read the article

• Prevent Linux from processing incoming ICMP Host unreachable packets

  - by bbc

  I have a test setup with one host on a network (10.1.0.0/16) talking via TCP to another one on another network (10.2.0.0/16) and a gateway in the middle. Sometimes, the TCP connection is lost and while scanning the trace (pcap), I looks like it's because of just one ICMP Host unreachable message sent by the gateway to 10.1.0.1 at some point. 10.1.0.1 then sends a TCP RST to 10.2.0.1. In my opinion, the gateway (pfSense) is broken or not configured correctly but anyway, for testing purposes, I'd like to block this kind of ICMP on the host (10.1.0.1) before it has an influence on my TCP connection (or does it? I'm not even sure). I've tried iptables: iptables -I INPUT -i eth0 -p icmp --icmp-type host-unreachable -j DROP but while it does a good job at preventing userpace applications like ping from receiving these ICMP messages, my TCP connection still comes to an end when the alleged "killer ICMP packet" is sent by the gateway. Am I right about how it is processed? If yes, then what can I do to achieve my goal?

  Read the article

• Finding bluetooth link key in Win7, to double pair a device on dualboot computer

  - by Ilari Kajaste

  How can I dig up the bluetooth link key for a paired device in Win7? Is this something that is dependent on the bluetooth stack I'm using (Toshiba), or is there a generic place to store these in Win7? Note: I'm not talking about the six-digit code usually typed by the user during pairing - that is worthless since it's discarded after pairing process. What I mean is the 128-bit link key that the devices exchange during pairing, and use thereafter to encrypt all their bluetooth traffic. Background: I dualboot Win7 / Ubuntu on my laptop, and I would like to have my phone paired to both OS's. Since the dualbooting computer has only one bluetooth adapter and thus only one bluetooth address, I cannot do two pairings to the phone, since on the second pairing (windows) the phone just replaces the previous pairing (linux) to the same bluetooth address. A thread on Ubuntu forums pointed me to what I have to do - pair first on linux, then on windows, and then replace the link key on linux side with the one windows negotiated. I can find the linux side pairing key from /var/lib/bluetooth/[BD_ADDR]/linkkeys - no problems there. However, on windows side I can't find the key. According to the forum post, on windows side the key should be in SYSTEM\ControlSet002\services\BTHPORT\Parameters\Keys\[BD_ADDR] but while that registry key does exist, it has no subkeys. (And a similar registry path in ControlSet001 didn't have any subkeys either.) One thing I've been instructed to do is to capture all events during pairing with Sysinternals Process Monitor. I did this, but I haven't been able to find any useful information from the captured events, not even by exporting the data to a huge XML and grepping that with the BD_ADDRs (with or without colons). So how could I find the link key for a paired device in Win7? Some reference information: Wikipedia: Bluetooth, Security Now: Bluetooth security

  Read the article

• Windows Server 2008 IIS Random disconnect

  - by d123

  I am having a bit of a quirk with my IIS server. I'm running my IIS with 2 sets of IPs configured, one in the 192 range and the other in 172 range. I then have multiple apps which will talk to this server for information. Server has no AV or firewalls configured. I noticed that my apps when talking to the server on the 172 range, at random intervals, the server would just not respond. My apps would then disconnect and just try again, and every thing would be fine. This doesn't happen on the 192 range. So what I did is on a Linux box I did a watch command and to wget a file every half second on the 172 and 192 IPs. I noticed the same issue, every once in awhile wget on the 172 range would not get through, but there is no issues at all on 192. Thus I went around to Wireshark and did a dump. This is the last 3 packets, no other packets were received. 7010 100.871877 200.100.30.7 172.0.0.1 TCP 59619 http [ACK] Seq=140 Ack=85242 Win=64128 Len=0 TSV=1072818795 TSER=1660246133 7011 100.872238 200.100.30.7 172.0.0.1 TCP 59619 http [FIN, ACK] Seq=140 Ack=85242 Win=64128 Len=0 TSV=1072818796 TSER=1660246133 7013 100.873081 200.100.30.7 172.0.0.1 TCP 59619 http [ACK] Seq=141 Ack=85243 Win=64128 Len=0 TSV=1072818796 TSER=1660246133 So this is my issue, there is a random disconnect every once in awhile. The server doesn't receive the next SYN packet. HELP?

  Read the article

• Basics about file/folder permissions on Win 7

  - by Altar

  Hi. Under Win XP I never touched the permissions of a file/folder. I was happy with the way it worked. But recently I installed Windows 7 on a drive that previously hosted Windows XP. Now, some programs do not have 'read' and/or 'write' access to their own folders - and I am not talking about system folders like 'Program Files' but normal folders like 'C:\my data\my own folder\program folder'. I see that for folders created under Win XP I have some user groups that do not exist for 'normal' folders (folders created by me recently under Windows 7). For example, for the Win XP folder I have: Creator owner System Account unknown(S-1-5-21 blablabla... Admins Users For Win7 folders I have: Authenticated users System Admins Users How should I proceed? Should I give the right to the "Users" account to write to XP folders? Should I make the old folders (the XP folders) to have the same groups of users as the normal (Win7) ones by adding the "Authenticated users" account to those folders? Should I delete the "Account unknown" account from my system? (In this case, how?) Many thanks.

  Read the article

• Are there any other causes of this error that are NOT related to initial setup?

  - by LordScree

  I'm trying to diagnose an issue at a customer site. They are receiving the following error: A network-related or instance-specific error occurred while establishing a connection to SQL Server I've seen this a few times, but only during the initial setup - it's often caused by one of the following: The database server is turned off The network connection between the database server and the application is closed or somehow blocked (e.g. a firewall) The SQL Server instance is not set up to receive remote connections from the application server (e.g. TCP is turned off, remote connections are disabled, or the "SQL Server Browser" service is stopped/disabled) However, if I assume that no configuration changes have been made, I'm trying to postulate on what the reason might be for getting this error at a random point after the initial setup. My initial thought is: SQL Server machine has run out of resources (e.g. RAM) and is unable to accept new requests from the application server Is this a valid theory? What other possible causes are there of this error that are not related to the initial setup of the server / application connection? Or is it simply impossible that this error could occur without a configuration change having been made (either on the SQL Server side, application side, or somewhere in-between (network))? NOTE: I believe this question differs from the plethora of questions related to this error message because the application and server have been talking to each other quite happily until now (most, if not all, other questions seem to relate to initial setup).

  Read the article

• Kindle (client) for Mac--text search or highlighting/notes?

  - by doug

  just so we're clear, i'm talking about the client/software version here--ie, that you install on your Mac or PC--not the device. The Kindle client was recently released for the Mac. I downloaded it and bought a couple of Kindle-edition books to view on this client. Astonishingly, two features i consider to be more or less essential to any ebook reader are missing in the Kindle client, either that, or i can't find them: (i) text searching; and (ii) highlighting text. First, does anyone know how to access the search feature? I'm aware of the "Go To" button at the top middle of the reader window--the options in that menu when you click the button are: "Cover", "Table of Contents", "Beginning" and "Location." "Location" requires that you type in an integer (but it doesn't correspond to page number--e.g., typing "167" brought me to the table of contents), not a search term. Second, there's a button on the upper right-hand corner of the window "Show Notes and Marks" yet i can't find any way to highlight text. The only kind of "note" or "mark" i have been able to record is to "bookmark" a page by clicking the "bookmark" button also at the top of the window.

  Read the article

• Remote Desktop Services Licensing - Does server have to have a RDS role?

  - by transistor1

  I recently set up a "micro" size Windows 2008 Datacenter server on Amazon AWS. My small group needs several concurrent RDS users to be able to access the machine. Without installing the "Remote Desktop Server" role, it allows 2 concurrent connections. I read on MS' website that in order to set up multiple users, we needed to install the RDS role. I did so, but now the application we are trying to share is running much slower than it was before. Prior to the role installation, it was taking about 5 seconds to open; now it is taking a few minutes to open -- without any other users logged on except me. My assumption is that the RDS role may be too much for this micro instance to handle, and currently, changing to another size instance is not an option (it may be possible later if we were to receive enough funding). This leads me to the following questions: 1) Is it a sensible assessment to assume that it is the RDS role is slowing things down, or are there other things that I could look at to speed it up? We are talking about a machine with ~600MB of memory. 2) If I revert back to the pre-RDS role, is there any legitimate way (in terms of purchasing RDS licenses) to get more than 2 concurrent desktops? I did read this, and am not questioning that the answerer is knowlegeable; but someone else may have some other experience. I am also making it clear that we want to do this in a legitimate way. Thanks in advance for any assistance that can be provided! EDIT: if it is helpful in answering the question, the application in question is a Lotus Approach database. Also, I am asking this from a technical perspective: not a legal one. I want to know if it is possible to install valid licenses without the RDS role.

  Read the article

• Is there any functional-like unix shell?

  - by Caruccio

  I'm (really) newbie to functional programming (in fact only had contact with it using python) but seems to be a good approach for some list-intensive tasks in a shell environment. I'd love to do something like this: $ [ git clone $host/$repo for repo in repo1 repo2 repo3 ] Is there any Unix shell with these kind of feature? Or maybe some feature to allow easy shell access (commands, env/vars, readline, etc...) from within python (the idea is to use python's interactive interpreter as a replacement to bash). EDIT: Maybe a comparative example would clarify. Let's say I have a list composed of dir/file: $ FILES=( build/project.rpm build/project.src.rpm ) And I want to do a really simple task: copy all files to dist/ AND install it in the system (it's part of a build process): Using bash: $ cp ${files[*]} dist/ $ cd dist && rpm -Uvh $(for f in ${files[*]}; do basename $f; done)) Using a "pythonic shell" approach (caution: this is imaginary code): $ cp [ os.path.join('dist', os.path.basename(file)) for file in FILES ] 'dist' Can you see the difference ? THAT is what i'm talking about. How can not exits a shell with these kind of stuff build-in yet? It's a real pain to handle lists in shell, even its being a so common task: list of files, list of PIDs, list of everything. And a really, really, important point: using syntax/tools/features everybody already knows: sh and python. IPython seams to be on a good direction, but it's bloated: if var name starts with '$', it does this, if '$$' it does that. It's syntax is not "natural", so many rules and "workarounds" ([ ln.upper() for ln in !ls ] -- syntax error)

  Read the article

• locale: What is the LANGUAGE variable used for? (and when?)

  - by seya

  I am trying to understand the locales used in Linux. On my Ubuntu 11.10 system locale puts out the following: LANG=en_DK.UTF-8 LANGUAGE=en_GB:en LC_CTYPE=en_GB.UTF-8 LC_NUMERIC="en_DK.UTF-8" LC_TIME="en_DK.UTF-8" LC_COLLATE=en_GB.UTF-8 LC_MONETARY="en_DK.UTF-8" LC_MESSAGES=en_GB.UTF-8 LC_PAPER="en_DK.UTF-8" LC_NAME="en_DK.UTF-8" LC_ADDRESS="en_DK.UTF-8" LC_TELEPHONE="en_DK.UTF-8" LC_MEASUREMENT="en_DK.UTF-8" LC_IDENTIFICATION="en_DK.UTF-8" LC_ALL= (en_dk is for using international day format, continental European number formatting (1.234,56) etc.) I think I understand what the LC_* family does, that LANG is the fallback if one of them is not set and that LC_ALL sets all of the LC_* variables to its value. What I don't know yet, is what LANGUAGE is used for. The notation en_GB:en reminds me of the Accept-Language HTTP header. With the settings above it would mean, British English is used, if a translation for it exists. Otherwise any existing English translation (en_US, en_AU, ..., whatever) would be used. Am I right so far? Also what programs actually obey the LANGUAGE setting? In how far is it different from LC_MESSAGES? Unfortunately, man locale only documents the LC_* family. And searching the web for 'linux locale LANGUAGE' or similar is a mute point. (Of course language is a word often used when talking about locales, and it may also be shown just in the output of locale without being discussed). Does anybody of you can help me out there?

  Read the article

• How can I disable flashing icons on Windows 7 taskbar?

  - by Jebego

  I set my Windows 7 taskbar to auto-hide. However, sometimes when a program changes or something new happens in a program, the taskbar will show its self, and its respective taskbar icon will begin flashing orange. Here's what I'm talking about: To make the taskbar hide again, I have click on the program before I can go back to what I was doing. Anyways, I personally find this very annoying, and would love to find a way to either: Prevent the taskbar from having such alerts. Prevent the taskbar from showing its self when it has such alerts. I've searched around quite a bit, and really only found answers to this for XP. I've also found another Stack Exchange Question looking for the same thing for Windows 7. However, none of the answers to the question were really what I'm looking for. I'm not looking to hide the taskbar, or control the number of flashes. However, this answer seems to be what I'm looking for, so I downloaded and tried out the program. It works perfectly, other than the fact that the start menu icon is always shown, regardless of the taskbar being set to auto-hide. So, any ideas on how to fix this problem?

  Read the article

• Is it possible to theme Windows 7 like XP?

  - by LonelyPixel

  Everybody seems to mean the window frame colour and a set of desktop background pictures when they're talking about Windows 7 themes. Does anybody remember what themes used to be in Windows XP? You could actually alter the appearance of the window frame, how close buttons and menu popups looked, put a texture on the taskbar, all those funny things. Microsoft themselves have published a number of XP themes over time (Luna on XP, Royal on MCE2005, Zune later on). I don't say that most of those and the numerous third party XP themes were beautiful or even practical. But it was possible to create something nice. In Windows 7 (I suppress the existence of Windows Vista...) such a theme could well be used to increase readability to a level before Aero Glass again. I mean, you cannot really tell whether a Glass window is active or not. I've been tweaking the colours and transparency levels a lot recently but the only safe indicator is the close button: it's red when the window is active, it's colourless otherwise. Then there's the window title. It is always painted black, on however dark a background. Again, regardless of whether the window is active or not. Turning off Aero is not an option in Windows 7 anymore. Classic design looks just ugly there. It already wasn't exactly looking good on XP with the wide start menu. So, can we increase the readability of the Windows 7 UI with themes like in XP or didn't Microsoft learn a thing since the Windows 7 Preview "Vista"?

  Read the article

• Fake demostration software for command line

  - by Joe

  I'm looking for some software that would be useful for giving demonstrations. I regularly have to show the effects of scrips ect to classes while talking about their effects, and equaly regularly I have finger trouble and have to rewrite various commands - wasting class time and general energy. I'd like to be able to record a sequence of commands in advance, and then play them back at the speed of my choosing. So I might have a file that containes the commands: echo "hello world!" ls ls -l ls -l | sort I'd like to be able to play these commands back by typing similar ones in. So I'd have a blinking command prompt and if I typed 'echo "hxxx' the command prompt would read home$echo "hell and if I typed any other letters the terminal would fill up with the remainder of the command until I press enter, when it executes the command. The point is that even if I screw up the command when typing it, the command that I'd prepared in advance would be executed. My question is - does similar software exist for giving demonstrations? or even, is this an easy thing to script up...? EDIT - two quick things first of all I'm on osx - but it would be nice to get a general solution for other people who arrive here from google. and second a lot of the comments/answers are concentrating on, in effect, making it fast and easy to enter long commands by means of hotkeys and the like. Actually I'd like it to at least look like I'm typing live - that's why I put in the bit about the one-to-one keymapping, but I don't think I explained that quite as well as I could have...

  Read the article

• Why still use JPG compression? [closed]

  - by Torben Gundtofte-Bruun

  Back when the JPG image format was introduced, it made a lot of sense to reduce the file size, even accepting a loss in image quality, because files were being downloaded over a slow and expensive modem connection. In today's world, file size is no longer a concern, at least not regarding JPG where it seems silly to save 45kB on a photo. But my image editing apps still prompt me for the desired compression level when I save a file. Does it still make sense to go with the default 85? Why should I not crank it up to 100 for all files? Update based on comments: For web work, I might use PNG instead. But every smartphone and camera produces JPG files. The question arises when I save these edits. Audience is my own harddisk. We're talking photos, 2-5MB apiece. Chroma, subsampling, DCT - sorry, never heard of it. I'm a home user, not Photoshop guru. For the record, I use Paint Shop Pro on Win, and Gimp on Linux.

  Read the article

• Zero downtime deployment (Tomcat), Nginx or HAProxy, behind hardware LB - how to "starve" old server?

  - by alexeypro

  Currently we have the following setup. Hardware Load Balancer (LB) Box A running Tomcat on 8080 (TA) Box B running Tomcat on 8080 (TB) TA and TB are running behind LB. For now it's pretty complicated and manual job to take Box A or Box B out of LB to do the zero downtime deployment. I am thinking to do something like this: Hardware Load Balancer (LB) Box A running Nginx on 8080 (NA) Box A running Tomcat on 8081 (TA1) Box A running Tomcat on 8082 (TA2) Box B running Nginx on 8080 (NB) Box B running Tomcat on 8081 (TB1) Box B running Tomcat on 8082 (TB2) Basically LB will be directing traffic between NA and NB now. On each of Nginx's we'll have TA1, TA2 and TB1, TB2 configured as upstream servers. Once one of the upstreams's healthcheck page is unresponsive (shutdown) the traffic goes to another one (HttpHealthcheckModule module on Nginx). So the deploy process is simple. Say, TA1 is active with version 0.1 of the app. Healthcheck on TA1 is OK. We start TA2 with Healthcheck on it as ERROR. So Nginx is not talking to it. We deploy app version 0.2 to TA2. Make sure it works. Now, we switch the Healthcheck on TA2 to OK, switch Healthcheck to TA1 to ERROR. Nginx will start serving TA2, and will remove TA1 out of rotation. Done! And now same with the other box. While it sounds all cool and nice, how do we "starve" the Nginx? Say we have pending connections, some users on TA1. If we just turn it off, sessions will break (we have cookie-based sessions). Not good. Any way to starve traffic to one of the upstream servers with Nginx? Thanks!

  Read the article

• Let varnish send old data from cache while it's fetching a new one?

  - by mark

  I'm caching dynamically generated pages (PHP-FPM, NGINX) and have varnish in front of them, this works very well. However, once the cache timeout is reached, I see this: new client requests page varnish recognizes the cache timeout client waits varnish fetches new page from backend varnish delivers new page to the client (and has page cached, too, for the next request which gets it instantly) What I would like to do is: client requests page varnish recognizes the timeout varnish delivers old page to the client varnish fetches new page from backend and puts it into the cache In my case it's not site where outdated information is such a big problem, especially not when we're talking about cache timeout from a few minutes. However, I don't want punish user to wait in line and rather deliver something immediate. Is that possible in some way? To illustrate, here's a sample output of running siege 5 minutes against my server which was configured to cache for one minute: HTTP/1.1,200, 1.97, 12710,/,1,2013-06-24 00:21:06 ... HTTP/1.1,200, 1.88, 12710,/,1,2013-06-24 00:21:20 ... HTTP/1.1,200, 1.93, 12710,/,1,2013-06-24 00:22:08 ... HTTP/1.1,200, 1.89, 12710,/,1,2013-06-24 00:22:22 ... HTTP/1.1,200, 1.94, 12710,/,1,2013-06-24 00:23:10 ... HTTP/1.1,200, 1.91, 12709,/,1,2013-06-24 00:23:23 ... HTTP/1.1,200, 1.93, 12710,/,1,2013-06-24 00:24:12 ... I left out the hundreds of requests running in 0.02 or so. But it still concerns me that there are going to be users having to wait almost 2 seconds for their raw HTML. Can't we do any better here? (I came across Varnish send while cache , it sounded similar but not exactly what I'm trying to do.)

  Read the article

• What are the practical differences between an IP address and a server?

  - by JMC Creative

  My understanding of IPs and other DNS-type server-related issues really falls short (read: exteme noob). I know a dedicated server would increase speed. What, if any, difference in speed would a dedicated IP make? Am I correct in understanding the Best Practices from Yahoo that I could use the second IP to serve up some content, which would increase the number of parallel downloads for the user? Or are both IPs (purchase from same hosting account) going to point to the same server? Or how does it work? Are there other optimization things I should be aware of when thinking of purchasing a dedicated IP? Clarification I am talking about the speed of serving the webpages, i.e. the speed of my website. Yes, I know that IP and server are completely different, not even opposites, just different. But this, indeed, is my question! The Question Reformulated: Will having a second (dedicated) IP on my website speed up the time that it will load and display for the user? Or does that have nothing at all to do with IP, and is only a server issue? I'm sorry if this is still unclear. This is a real question though, I may just not be wording it well.

  Read the article

• Identifying test machines in analytics logs

  - by RTigger

  We're just beginning to add analytics to our SaaS application, to begin (among other things) billing clients based on usage. The problem we're running into is there's a few circumstances where our support team will simulate a log in into production to try to reproduce reported issues with a client's configuration. When they log in, an entry will be made into our analytics logs that their specific account has logged in, which we use to calculate billing. A few ideas we had to solve this: 1) We log IP addresses as well as machine keys for each PC that logs in - we could filter out known IP addresses and/or machine keys belonging to support. The drawback is we have to maintain a list of keys / addresses manually. 2) If support (or anyone else internal) runs our application in debug mode (as opposed to release), it will not report analytics. This is fine, as long as support / anyone else remembers to switch to debug mode. 3) Include some sort of reg key / similar setting required to be set when configuring a production system in order to send analytics. Again, fine, as long as our infrastructure team remembers to set the reg key or setting. All of these approaches require some sort of human involvement, which we all know can be iffy at best. Has anyone run into a similar situation? Is there an automated approach to this problem? (PS Of course, we shouldn't be testing in production, but there are a few one-off instances with customer set up that we can't reproduce without logging in as them in production. This is the only time we do so, and this is the case I'm talking about in this question.)

  Read the article

• Troubleshooting DTCPing Errors

  - by JimmyP

  So I am running DTC ping between 2 machines on our network and am getting the following error ++++++++++++++++++++++++++++++++++++++++++++++ DTCping 1.9 Report for WEB2 ++++++++++++++++++++++++++++++++++++++++++++++ RPC server is ready ++++++++++++Validating Remote Computer Name++++++++++++ 03-03, 13:39:45.099-->Start DTC connection test Name Resolution: internal-->10.20.3.236-->internal.something 03-03, 13:39:45.114-->Start RPC test (WEB2-->internal) Problem:fail to invoke remote RPC method Error(0x6BA) at dtcping.cpp @303 -->RPC pinging exception -->1722(The RPC server is unavailable.) RPC test failed I have also run RPC ping where I get what I beleive is the same error: C:\Program Files\Windows Resource Kits\Tools>rpcping -s internal Exception 1722 (0x000006BA) Number of records is: 4 ProcessID is 5876 System Time is: 3/3/2011 2:44:12:822 Generating component is 8 Status is 1722 Detection location is 323 Flags is 0 NumberOfParameters is 0 ProcessID is 5876 System Time is: 3/3/2011 2:44:12:822 Generating component is 8 Status is 1237 Detection location is 313 Flags is 0 NumberOfParameters is 0 ProcessID is 5876 System Time is: 3/3/2011 2:44:12:822 Generating component is 8 Status is 10060 Detection location is 311 Flags is 0 NumberOfParameters is 3 Long val: 135 Pointer val: 0 Pointer val: 0 ProcessID is 5876 System Time is: 3/3/2011 2:44:12:822 Generating component is 8 Status is 10060 Detection location is 318 Flags is 0 NumberOfParameters is 0 I'm pretty sure that the exception number 1722 is the key but I can't find any info about it. There may be a firewall with ports that need opening between the machines which I am checking with our sys admins now. But I can do a regular ping between the machines. Other than that I am reading a lot of articles talking about OS services and components I know nothing about and am having trouble finding any info on. Can anyone shed any light on this? FYI the machine is running Windows Server 2003 RS SP2.

  Read the article

< Previous Page | 89 90 91 92 93 94 95 96 97 98 99 100  | Next Page >