Search Results

Search found 18964 results on 759 pages for 'network balancing'.

Page 3/759 | < Previous Page | 1 2 3 4 5 6 7 8 9 10 11 12  | Next Page >

  • Utilize two gateways on the same network same interface with load balancing

    - by RushPL
    My setup is two ISPs on a single interface and single network. I can either set my default gateway to 192.168.0.1 or 192.168.1.250 and either work. My desire is to utilize both of them with some load balancing. I have tried to follow the advice given in here http://serverfault.com/a/96586 #!/bin/sh ip route show table main | grep -Ev '^default' \ | while read ROUTE ; do ip route add table ISP1 $ROUTE done ip route add default via 192.168.1.250 table ISP1 ip route add default via 192.168.0.1 table ISP2 iptables -t mangle -A PREROUTING -j CONNMARK --restore-mark iptables -t mangle -A PREROUTING -m mark ! --mark 0 -j ACCEPT iptables -t mangle -A PREROUTING -j MARK --set-mark 10 iptables -t mangle -A PREROUTING -m statistic --mode random --probability 0.5 -j MARK --set-mark 20 iptables -t mangle -A PREROUTING -j CONNMARK --save-mark Now then I do "traceroute somehost" repeatedly I can only get route through my default route which is 192.168.1.250. Shouldn't the packets change routes in a random manner? How to debug it?

    Read the article

  • apache, shibboleth, load balancing aliase, ssl

    - by Nikolaidis Fotis
    Good morning folks Could you give me a bit of help with the following problem ? I have a dns load balancing mechanism and an alias (hostAlias) which may point to host01, or host02 I want to configure apache and shibboleth to work with that alias. What happens is ... User types : https://hostAlias (it points to host01) apache host01 : redirect to shibboleth shibboleth host01 : redirect to **https://hostAlias.cern.ch/Shibboleth.sso/ADFS** Now, there are two cases. Either this time hostAlias will point again to host01 , or it will point to host02. If it points to host02, host01 will not get the anwser and the authentication fails. Also, about ssl certificates, I guess that each host will need its own certificate. right ? Should I need a certificate with DNS aliases ? Thanks in advance !

    Read the article

  • Sticky Load Balancing with AWS

    - by John Wheal
    I have just setup a load balancer with AWS for a few instances as search engine crawlers were bringing down the site (it has millions of pages). Parts of the site allow you to login so I selected: Enable Application Generated Cookie Stickiness and everything works fine. I now wonder how this will effect my SEO and the crawlers. As I selected sticky load balancing does this mean that a crawler will be stuck on one server and therefore defeat the point in the load balancer? Any recommendations will be appreciated.

    Read the article

  • Amazon EC2 Elastic Load Balancing - strategy for zero downtime server restart

    - by Yoga
    I have 5 web servers (Apache/mod_perl) behind Amazon EC2 Elastic Load Balancing, when I deploy codes to the web servers, I am doing this.. For each machine, shutdown the Apache Update the code Start over the server and proceed to the next server I think when my server is shutdown, ELB will not distribute request to my server, but how about the request still serving? I think a better approach is Stop accepting new request from ELB Sleep for sometimes, shutdown web server only if all requests are responded Update the codes Start the server again But how to perform (1) and (2) from my local sever? Do I need to use AWS API? or other easy way to do it? Thanks.

    Read the article

  • A really unique case of Load Balancing

    - by Shamshun
    I have an internet connection which has limited up/down bandwidth per IP address. What I want to do is to get multiple IP addresses on a "single" LAN interface, and use a load balancer to distribute traffic through them. I was successful at getting 100 ip addresses on a single interface. my problem is that Linux and Windows use the first ip address of an interface by default, so my bandwidth is not increased. I would appreciate if someone tells me what Load Balancing software has the ability to distribute load between multiple IPs on a single interface. I have tried to do so on both Win7 and Backtrack-LinuxR2

    Read the article

  • How do I ban a wifi network in Network Manager?

    - by Chris Conway
    My wifi connection drops sometimes and, for some reason, Network Manager attempts to connect to my neighbor's network, which requires a password that I don't know. The network in question is not listed in the "Edit Connections..." dialog and I can find no reference to it in any configuration file, but still the password dialog pops up every time my main connection drops. Is there a way to blacklist a wireless network so that the Network Manager will never attempt to connect to it? Or, equivalently, how can I remove the configuration data that causes the Network Manager to attempt to connect to this particular network?

    Read the article

  • Adding network share as the system account breaks Win7 backup to network

    - by ChrisBenn
    (On Win7 Ultimate x64 SP1) So I've been using Win7 backup to \192.168.0.100\Backup\main-desktop\ for awhile without issue. Yesterday I tried to setup crashplan to synchronize my dropbox folder and a network share. I then found out that crashplan, as it runs under the system account, can't see my user mapped drives. So I created a startup script net use O: \192.168.0.100\Documents /USER:192.168.0.100\username password and set it to run, on startup, after the network interface is up, as the system account. (the username & password are the same for the net use script above, the locally logged in user, and the explicit username/password entered in windows backup). I woke up this morning to find error flags from the windows backup and get "Network location cannot be used" (0x800704B3). If I disable the startup task & reboot then windows backup works fine. I'm not sure why having a mapped drive for another user is killing windows backup (same server, different folder). I can work around the issue by using another program to synchronize the two folders, but I'm completely in the dark as to why this happens (and it's 100% repeatable). Uninstalling the crashplan client doesn't change anything - it's the net use run under the system account that breaks win7 backup (to a network location).

    Read the article

  • Amazon CloudFront and EC2: Global Load Balancing

    - by Matt Rogish
    We have an app that is going to store and serve up a decent amount of data in S3 to a global audience where latency should be minimized. So, we've been doing tests with Amazon CloudFront and have seen favorable results. However, we need a thin middleware layer (to do security etc.) and we'd like to put that in EC2. Due to security restrictions, this middleware layer will do the file streaming from S3/CloudFront: S3/CloudFront - EC2 - Clients We can geographically distribute the EC2 nodes (US East/West, and Ireland) but the problem is that a client in the EU would hit our US server and be fed data from there, thus rendering much of the performance benefit of CloudFront moot. I've been digging through the EC2 docs but I can't find a built-in way to get a geographically distributed version of EC2 a la CloudFront. Elastic Load Balancing sounds like the way to go, but I can't seem to find a way with that to direct based on routing... Preferably, we'd like to keep the amount of stuff outside of EC2/S3/etc. to a minimum (for obvious reasons). Any ideas how to do that within the EC2/S3 framework? DNS/routing tricks? Thanks!

    Read the article

  • Cisco Catalyst 3550 + Alteon 184 Load-Balancing Issues...

    - by upkels
    I have just deployed a couple Cisco Catalyst 3550 switches, and a couple Alteon 184 Web Switches for load-balancing. I can ping all RIPs and VIPs to/from the Alteon. Topology Before: (server) <- (Alteon) <- (Internet) Topology Now: (server) <- (3550) <- Alteon <- (Internet) Cisco Port Configuration (Alteon Uplink Port): description LB_1_PORT_9_PRIMARY switchport access vlan 10 switchport mode access switchport nonegotiate speed 100 duplex full Alteon Port 9 Configuration (VLAN 10 WAN): >> Main# /c/port 9/cur Current Port 9 configuration: enabled pref fast, backup gig, PVID 10, BW Contract 1024 name UPLINK >> Main# /c/port 9/fast/cur Current Port 9 Fast link configuration: speed 100, mode full duplex, fctl none, auto off Cisco Configuration (Load-Balanced Servers Port): description LB_1_PORT_1_PRIMARY switchport access vlan 30 switchport mode access switchport nonegotiate speed 100 duplex full Alteon Port 1 Configuration (VLAN 30 LOAD-BALANCED LAN): >> Main# /c/port 1/cur Current Port 1 configuration: enabled pref fast, backup gig, PVID 30, BW Contract 1024 name LB_PORT_1 >> Main# /c/port 1/fast/cur Current Port 1 Fast link configuration: speed 100, mode full duplex, fctl both, auto on Each of my servers are on vlan 10 and 30, properly communicating. I have tried to turn on VLAN tagging on the Alteon, however it seems to cause all communications to stop working. When I tcpdump -i vlan30 on any of the webservers, I see normal ARP communications, and some STP communications, which may or may not be part of the problem: ... 15:00:51.035882 STP 802.1d, Config, Flags [none], bridge-id 801e.00:11:5c:62:fe:80.8041, length 42 15:00:51.493154 IP 10.1.1.254.33923 > 10.1.1.1.http: Flags [S], seq 707324510, win 8760, options [mss 1460], length 0 15:00:51.493336 IP 10.1.1.1.http > 10.1.1.254.33923: Flags [S.], seq 3981707623, ack 707324511, win 65535, options [mss 1460], len gth 0 15:00:51.493778 ARP, Request who-has 10.1.3.1 tell 10.1.3.254, length 46 etc... I'm not sure if I've provided enough information, so please let me know if any more is necessary. Thank you!

    Read the article

  • Apache, Tomcat and mod_jk for load balancing

    - by pHk
    Hi guys. I've set-up a basic Apache (2.2.x) and Tomcat (6.0.x) set-up using mod_jk for load balancing using the worker.properties file. Preliminary testing seems to show that this works relatively well, and it was quite easy to set-up. However; the fact that it was so easy to set-up has got me a little worried. We're dealing with 100 - 300 concurrent users using the same web application (deployed on 2 or 3 Tomcat instances). I have done a little Googling and looking around on here and there seems to be more than 1 way to accomplish this (one example on here used a balancer:// style URL, which I've never seen before in an Apache config). For example, one question I ask myself is how reliable the load detection on mod_jk really is (Busyness, Session, Request, etc). In your experience, does this set-up prove to be reliable in real world scenarios? Any pointers on improvements, pit falls or interesting literature/articles? I've worked with Apache before, but am in no way an expert. Thanks in advance.

    Read the article

  • Cisco Catalyst 3550 + Alteon 184 Load-Balancing Issues

    - by upkels
    I have just deployed a couple Cisco Catalyst 3550 switches, and a couple Alteon 184 Web Switches for load-balancing. I can ping all RIPs and VIPs to/from the Alteon. Topology Before: (server) <- (Alteon) <- (Internet) Topology Now: (server) <- (3550) <- Alteon <- (Internet) Cisco Port Configuration (Alteon Uplink Port): description LB_1_PORT_9_PRIMARY switchport access vlan 10 switchport mode access switchport nonegotiate speed 100 duplex full Alteon Port 9 Configuration (VLAN 10 WAN): >> Main# /c/port 9/cur Current Port 9 configuration: enabled pref fast, backup gig, PVID 10, BW Contract 1024 name UPLINK >> Main# /c/port 9/fast/cur Current Port 9 Fast link configuration: speed 100, mode full duplex, fctl none, auto off Cisco Configuration (Load-Balanced Servers Port): description LB_1_PORT_1_PRIMARY switchport access vlan 30 switchport mode access switchport nonegotiate speed 100 duplex full Alteon Port 1 Configuration (VLAN 30 LOAD-BALANCED LAN): >> Main# /c/port 1/cur Current Port 1 configuration: enabled pref fast, backup gig, PVID 30, BW Contract 1024 name LB_PORT_1 >> Main# /c/port 1/fast/cur Current Port 1 Fast link configuration: speed 100, mode full duplex, fctl both, auto on Each of my servers are on vlan 10 and 30, properly communicating. I have tried to turn on VLAN tagging on the Alteon, however it seems to cause all communications to stop working. When I tcpdump -i vlan30 on any of the webservers, I see normal ARP communications, and some STP communications, which may or may not be part of the problem: ... 15:00:51.035882 STP 802.1d, Config, Flags [none], bridge-id 801e.00:11:5c:62:fe:80.8041, length 42 15:00:51.493154 IP 10.1.1.254.33923 > 10.1.1.1.http: Flags [S], seq 707324510, win 8760, options [mss 1460], length 0 15:00:51.493336 IP 10.1.1.1.http > 10.1.1.254.33923: Flags [S.], seq 3981707623, ack 707324511, win 65535, options [mss 1460], len gth 0 15:00:51.493778 ARP, Request who-has 10.1.3.1 tell 10.1.3.254, length 46 etc... I'm not sure if I've provided enough information, so please let me know if any more is necessary. Thank you!

    Read the article

  • Nginx load balancing and maintaining URLs

    - by Steve Klabnik
    I'm trying to use nginx as a load balancer, and it's working great. One problem, though. The load balancing box is at 123.123.123.123, and the backend box is 456.456.456.456. So I have this config: upstream backend { server 456.456.456.456; } server { listen 80; server_name 123.123.123.123; access_log off; error_log off; location / { proxy_set_header Host $host; proxy_set_header X-Real-IP $remote_addr; proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; proxy_pass http://backend; } } This works great. I hit 123.123.123.123 in my browser, and the page comes up. But now the URL in the browser says http://456.456.456.456. Do I need to use a rewrite rule or something to keep the url correct? I don't want it to be different when going to different backed servers. None of the tutorials I've read have mentioned anything about this.

    Read the article

  • HAproxy roundrobin balancing does not appear to be distributing evently

    - by andrew
    Hello, I know that with loaded servers, roundrobin in HAproxy (1.4.4) does not evenly distribute, but my servers are currently getting NO traffic (test setup), and roundrobin balancing does www1,www1,www1,www1,www1,...www2,www2,www2,...,www1... I'm verifying this by having the script being run on each server cat /etc/HOSTNAME (slackware). I need to have it switch back and forth each time to test some session stuff (stored in shared memcached) but am having trouble getting it to switch between my two web servers on each request. global log 127.0.0.1 local0 warning maxconn 4096 chroot /usr/share/haproxy pidfile /var/run/haproxy.pid uid 99 gid 99 daemon defaults balance roundrobin fullconn 100 maxconn 4096 mode http option dontlognull option http-server-close option forwardfor option redispatch retries 3 timeout connect 5000 timeout client 20000 timeout server 60000 timeout queue 60000 stats enable stats uri /haproxy stats auth ***:*** frontend www *:80 log global acl is_upload hdr_dom(host) -i uploads.site.com acl is_api hdr_dom(host) -i api.site.com acl is_dev hdr_dom(host) -i dev.site.com acl is_apidev hdr_dom(host) -i apidev.site.com use_backend uploads.site.com if is_upload use_backend api.site.com if is_api use_backend dev.site.com if is_dev !is_apidev default_backend site.com backend site.com option httpchk HEAD /alive.php HTTP/1.1\r\nHost:site.com server www1 1.1.1.1:8080 weight 10 minconn 5 maxconn 25 check inter 2000 rise 2 fall 2 server www2 1.1.1.2:8080 weight 10 minconn 5 maxconn 25 check inter 2000 rise 2 fall 2 backend api.site.com option httpchk HEAD /alive.php HTTP/1.1\r\nHost:api.site.com server www1 1.1.1.1:8080 weight 10 minconn 5 maxconn 25 check inter 2000 rise 2 fall 2 server www2 1.1.1.2:8080 weight 10 minconn 5 maxconn 25 check inter 2000 rise 2 fall 2 backend dev.site.com option httpchk HEAD /alive.php HTTP/1.1\r\nHost:dev.site.com server www1 1.1.1.1:8080 weight 10 minconn 5 maxconn 25 check inter 2000 rise 2 fall 2 server www2 1.1.1.2:8080 weight 10 minconn 5 maxconn 25 check inter 2000 rise 2 fall 2 backend uploads.site.com option httpchk HEAD /alive.php HTTP/1.1\r\nHost:uploads.site.com server www1 1.1.1.1:8080 weight 10 minconn 5 maxconn 25 check inter 2000 rise 2 fall 2 server www2 1.1.1.2:8080 backup weight 10 minconn 5 maxconn 25 check inter 2000 rise 2 fall 2 So basically, I have some different back-ends (I've verified the ACLs are working), with the default option "roundrobin" selected. I've tried removing weights, removing the minconn/maxconn/fullconn attributes for all servers (not just the backend I'm testing), tried removing the ACLs, etc. I've been testing on dev.site.com BTW. Anyone see a reason why I can't get something like www1,www2,www1,www2,...? Also, this is one of my first questions on here, so please let me know if I left anything needed out of my post. Thanks!

    Read the article

  • Which network performance management software do you use?

    - by Jamie Keeling
    Hello, I am looking at the various options available for network performance management software, some of the solutions I've found so far are: Proprietary: HP - ProCurve Universal: SolarWinds - Orion Open Source: OpenNMS I am trying to discover the benefits of each package over the other and reasons as to why you would go for one (Such as size of the network, overall cost etc..). I'm curious as to which ones other people use and why? Each customer has their own needs and requirements and it would be great to hear some of yours. Thank you for your time.

    Read the article

  • Getting Xbox Live via a wired network with my laptop that has internet access wirelessly

    - by Alex Franco
    I'm running the latest version (as of yesterday anyways) of Ubuntu Desktop 64bit, but installed on my laptop if it makes a difference. I had Windows 7 preinstalled when i bought it and it worked fine with the wireless from my house and bridging the connection with a LAN to my xbox for Live. Now with Ubuntu I tried the same setup, but I'm unfamiliar with Ubuntu so I didn't get far. Best I got so far is wireless internet on my laptop and a wired connection to the xbox that continually connects and disconnects. Heres my network settings. if theres fields not included its because theyre empty on mine or theyre my MAC address or network password Wireless Network 1 settings: Connect Automatically: Checked. Available to all Users: Checked Wireless: SSID: Franco's Mode: Infrastructure MTU: Automatic IPv4 Settings: Method: Automatic (DHCP) IPv6 Settings: Method: Automatic Wired Network 1: Connect Automatically: Checked Available to all Users: Checked Wired: MTU: Automatic IPv4 Settings: Method: Automatic (DHCP) IPv6 Settings: Method: Automatic Any help would be greatly appreciated. EDIT: 6:26pm It seems to be staying connected now. Doing the Network test on my xbox it pickups the network, but cannot detect any PC. Restarting the Xbox, however, leaves my computer unable to connect bringing up the Wire Network disconnected 'blip' every minute or so again. Before I had restarted the Xbox it said "Connected 100 MB/s". Now it only says "connecting". I did have my computer and xbox on in this Wired Network Disconnected blip cycle for a long period of time so it may have finally connected, just without the ability to detect my laptop. I left for 2 hours or so in the middle of typing up the original question. I finished posting this when i got back and then tried to mess with it a bit again, in case youre wondering why i didnt include this before... I've said too much. Forgive my long-winded fingers :p

    Read the article

  • Load balancing with Cisco router

    - by you8301083
    I have a Cisco router with two bonded T1's which are setup as a VPN to the main office. We need more bandwidth but can't get other connections (or it's too costly), so I would like to have a dsl connection installed. This DSL connection will run over a VPN to the same main office, but it won't be bonded with the T1's - so it won't act as a single connection. Since the three circuits won't act as a single connection (basically would be two connections 2 T1's + 1 DSL) we would have to split the network in half - but I don't want to do that. Instead, would it be possible to send all HTTP/HTTPS over the DSL connection but send all mission critical data (such as voice/active directory) over the T1's? I basically want to send specific ports over DSL and everything else over the T1's without separating half of the users traffic over the DSL and the rest over the T1's.

    Read the article

  • Requests per second slower when using nginx for load balancing

    - by Ed Eliot
    I've set up nginx as a load balancer that reverse proxies requests to 2 Apache servers. I've benchmarked the setup with ab and am getting approx 35 requests per second with requests distributed between the 2 backend servers (not using ip_hash). What is confusing me is that if I query either of the backend servers directly via ab I get around 50 requests per second. I've experimented with a number of different values in ab the most common being 1000 requests with 100 concurrent connections. Any idea why traffic distributed across 2 servers would result in fewer requests per second than hitting either directly? Additional info: I've experimented with worker_processes values of between 1 and 8, worker_connections between 1024 and 8092 and have also tried keepalive 0 and 65. My main conf currently looks like this: user www-data; worker_processes 1; error_log /var/log/nginx/error.log; pid /var/run/nginx.pid; worker_rlimit_nofile 8192; events { worker_connections 2048; use epoll; } http { include /etc/nginx/mime.types; sendfile on; keepalive_timeout 0; tcp_nodelay on; gzip on; gzip_disable "MSIE [1-6]\.(?!.*SV1)"; include /etc/nginx/conf.d/*.conf; include /etc/nginx/sites-enabled/*; } I've got one virtual host (in sites available) that redirects everything under / to 2 backends across a local network.

    Read the article

  • Network printing -- setting up and printing over network using Mac OS X

    - by crippledlambda
    I've connected a USB printer to a Mac OS X (10.4) machine -- this computer is connected to a network with a registered IP address -- and enabled sharing through the System Preferences setting (under "Shared" and "Print & Fax". Is this printer shared through the Bonjour or CUPS protocol (does it matter that I know)? How do I sent a print job to this computer (also from a 10.4 OS X machine)? It does not show up in the list of printers in the Default Network when I go to "add printer"; do I need to do anything else (on either machine)? Thanks much!

    Read the article

  • Do i need a dedicated server for load balancing?

    - by Ben
    I'm completely new to the concept of load balancing so i hope this question isn't a "stupid question" because i've been searching around and im having a hard time understanding this. So to my understanding, in order to load balance, i need a separate machine with an ip address i can direct all traffic to. I initially thought i needed to rent 3 dedicated servers, one for load balancing and the other two as backend servers. Would a dedicated server be too much for a load balancer or do hosting companies have special types of computers for that process? Then i read somewhere else that i can install a load balance software in both of the two servers and configure it in a way that doesn't require me to rent another machine/dedicated server for load balancing. So im a bit confuse on how to actually implement a load balancer and whether or not i need a dedicated server for the sole purpose of acting as a load balancing machine. Also, i was recommended to use HAproxy so i'll be heading that direction for load balancing.

    Read the article

  • Conceptually how does load-balancing on the EJB tier work in Glassfish/any ejb container

    - by Benju
    I am wondering conceptually how load-balancing works on the EJB-level (not web session replication) with Java EE containers like Glassfish. From what I have gleaned your remote interface is a proxy that delegates your call to one of many servers you may have in an environment. If things fail are they supposed to be able to "finish" on another server? I want to understand the basic theory behind this load balancing, why is it better than a bunch of servers all running a plain web application with session affinity on a load-balancer?

    Read the article

  • Neural Network Always Produces Same/Similar Outputs for Any Input

    - by l33tnerd
    I have a problem where I am trying to create a neural network for Tic-Tac-Toe. However, for some reason, training the neural network causes it to produce nearly the same output for any given input. I did take a look at Artificial neural networks benchmark, but my network implementation is built for neurons with the same activation function for each neuron, i.e. no constant neurons. To make sure the problem wasn't just due to my choice of training set (1218 board states and moves generated by a genetic algorithm), I tried to train the network to reproduce XOR. The logistic activation function was used. Instead of using the derivative, I multiplied the error by output*(1-output) as some sources suggested that this was equivalent to using the derivative. I can put the Haskell source on HPaste, but it's a little embarrassing to look at. The network has 3 layers: the first layer has 2 inputs and 4 outputs, the second has 4 inputs and 1 output, and the third has 1 output. Increasing to 4 neurons in the second layer didn't help, and neither did increasing to 8 outputs in the first layer. I then calculated errors, network output, bias updates, and the weight updates by hand based on http://hebb.mit.edu/courses/9.641/2002/lectures/lecture04.pdf to make sure there wasn't an error in those parts of the code (there wasn't, but I will probably do it again just to make sure). Because I am using batch training, I did not multiply by x in equation (4) there. I am adding the weight change, though http://www.faqs.org/faqs/ai-faq/neural-nets/part2/section-2.html suggests to subtract it instead. The problem persisted, even in this simplified network. For example, these are the results after 500 epochs of batch training and of incremental training. Input |Target|Output (Batch) |Output(Incremental) [1.0,1.0]|[0.0] |[0.5003781562785173]|[0.5009731800870864] [1.0,0.0]|[1.0] |[0.5003740346965251]|[0.5006347214672715] [0.0,1.0]|[1.0] |[0.5003734471544522]|[0.500589332376345] [0.0,0.0]|[0.0] |[0.5003674110937019]|[0.500095157458231] Subtracting instead of adding produces the same problem, except everything is 0.99 something instead of 0.50 something. 5000 epochs produces the same result, except the batch-trained network returns exactly 0.5 for each case. (Heck, even 10,000 epochs didn't work for batch training.) Is there anything in general that could produce this behavior? Also, I looked at the intermediate errors for incremental training, and the although the inputs of the hidden/input layers varied, the error for the output neuron was always +/-0.12. For batch training, the errors were increasing, but extremely slowly and the errors were all extremely small (x10^-7). Different initial random weights and biases made no difference, either. Note that this is a school project, so hints/guides would be more helpful. Although reinventing the wheel and making my own network (in a language I don't know well!) was a horrible idea, I felt it would be more appropriate for a school project (so I know what's going on...in theory, at least. There doesn't seem to be a computer science teacher at my school). EDIT: Two layers, an input layer of 2 inputs to 8 outputs, and an output layer of 8 inputs to 1 output, produces much the same results: 0.5+/-0.2 (or so) for each training case. I'm also playing around with pyBrain, seeing if any network structure there will work. Edit 2: I am using a learning rate of 0.1. Sorry for forgetting about that. Edit 3: Pybrain's "trainUntilConvergence" doesn't get me a fully trained network, either, but 20000 epochs does, with 16 neurons in the hidden layer. 10000 epochs and 4 neurons, not so much, but close. So, in Haskell, with the input layer having 2 inputs & 2 outputs, hidden layer with 2 inputs and 8 outputs, and output layer with 8 inputs and 1 output...I get the same problem with 10000 epochs. And with 20000 epochs. Edit 4: I ran the network by hand again based on the MIT PDF above, and the values match, so the code should be correct unless I am misunderstanding those equations. Some of my source code is at http://hpaste.org/42453/neural_network__not_working; I'm working on cleaning my code somewhat and putting it in a Github (rather than a private Bitbucket) repository. All of the relevant source code is now at https://github.com/l33tnerd/hsann.

    Read the article

  • Best Solution for Load Balancing NFS File Access?

    - by DairyKnight
    I'm trying to find an optimum solution for accessing the NFS file share in my company. We have a central file server in North America and has 30GB~50GB of updated data everyday. And it's very slow for our Europe and Asia branches to access directly. Therefore, I'm trying to setup two replicate servers in those continents. I'm currently using rsync, but wonder if there exists a better solution acts more like a distributed RAID, which allows the user to transparently access the file whether synced or not. And user request will be dispatched to remote server if the file is not yet synced. I'm now looking into DRBD, but it seems not to have the functionality of auto-dispatching requests. Does anyone know if there's a better solution?

    Read the article

< Previous Page | 1 2 3 4 5 6 7 8 9 10 11 12  | Next Page >