Search Results

Search found 20797 results on 832 pages for 'twisted client'.

Page 7/832 | < Previous Page | 3 4 5 6 7 8 9 10 11 12 13 14  | Next Page >

  • How to use Cisco AnyConnect VPN Client?

    - by ktm5124
    I wrote a related question earlier, which is still unresolved. This question is much more specific. So I installed Cisco AnyConnect VPN Client on Snow Leopard. I connect to my work VPN. Once connected, I can't ping my work machine. I don't see any computers on the network. If the client were not running, I wouldn't believe myself to be connected to the VPN. Is there something that I am doing wrong? Do I have to route my network traffic through the tunnel manually? (ifconfig route comes to mind) Is the POST request that I am about to submit going to go through the tunnel created by my VPN? I guess the main question is: why do I feel so in the dark? Cisco says I am connected to my VPN, but for all I know it is invisible. N.B. I do have the up-to-date Cisco VPN Client: version 2.3.2016. I installed it about a week ago.

    Read the article

  • tmux: Suddenly, cannot horizontally split

    - by A__A__0
    As root, using a reasonably default .profile and .shrc and an empty tmux.conf, I am unable to split the window horizontally. There are a number of cases to consider so I'll list them clearly. Using the keybinding + empty configuration: nothing happens Using the keybinding + my configuration: a bell is generated, nothing else; occasionally, the split will appear and disappear immediately (maybe it always does this, but I'm connecting over ssh so it may not make it through) Using tmux split-window -h with any config: tmux immediately exits I've posted here in order the server and client verbose logs generated by tmux -v during the third case: server started, pid 9523 socket path /tmp/tmux-0/default new client 7 got 100 from client 7 got 101 from client 7 got 102 from client 7 got 103 from client 7 got 104 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 105 from client 7 got 106 from client 7 got 200 from client 7 cmdq 0x801c6e080: new-session (client 7) new term: xterm xterm override: XT xterm override: Ms ]52;%p1%s;%p2%s xterm override: Cs ]12;%p1%s xterm override: Cr ]112 xterm override: Ss [%p1%d q xterm override: Se [2 q new key Oo: 0x1021 (KP/) new key Oj: 0x1022 (KP*) new key Om: 0x1023 (KP-) new key Ow: 0x1024 (KP7) new key Ox: 0x1025 (KP8) new key Oy: 0x1026 (KP9) new key Ok: 0x1027 (KP+) new key Ot: 0x1028 (KP4) new key Ou: 0x1029 (KP5) new key Ov: 0x102a (KP6) new key Oq: 0x102b (KP1) new key Or: 0x102c (KP2) new key Os: 0x102d (KP3) new key OM: 0x102e (KPEnter) new key Op: 0x102f (KP0) new key On: 0x1030 (KP.) new key OA: 0x101d (Up) new key OB: 0x101e (Down) new key OC: 0x1020 (Right) new key OD: 0x101f (Left) new key [A: 0x101d (Up) new key [B: 0x101e (Down) new key [C: 0x1020 (Right) new key [D: 0x101f (Left) new key OH: 0x1018 (Home) new key OF: 0x1019 (End) new key [H: 0x1018 (Home) new key [F: 0x1019 (End) new key Oa: 0x501d (C-Up) new key Ob: 0x501e (C-Down) new key Oc: 0x5020 (C-Right) new key Od: 0x501f (C-Left) new key [a: 0x901d (S-Up) new key [b: 0x901e (S-Down) new key [c: 0x9020 (S-Right) new key [d: 0x901f (S-Left) new key [11^: 0x5002 (C-F1) new key [12^: 0x5003 (C-F2) new key [13^: 0x5004 (C-F3) new key [14^: 0x5005 (C-F4) new key [15^: 0x5006 (C-F5) new key [17^: 0x5007 (C-F6) new key [18^: 0x5008 (C-F7) new key [19^: 0x5009 (C-F8) new key [20^: 0x500a (C-F9) new key [21^: 0x500b (C-F10) new key [23^: 0x500c (C-F11) new key [24^: 0x500d (C-F12) new key [25^: 0x500e (C-F13) new key [26^: 0x500f (C-F14) new key [28^: 0x5010 (C-F15) new key [29^: 0x5011 (C-F16) new key [31^: 0x5012 (C-F17) new key [32^: 0x5013 (C-F18) new key [33^: 0x5014 (C-F19) new key [34^: 0x5015 (C-F20) new key [2^: 0x5016 (C-IC) new key [3^: 0x5017 (C-DC) new key [7^: 0x5018 (C-Home) new key [8^: 0x5019 (C-End) new key [6^: 0x501a (C-NPage) new key [5^: 0x501b (C-PPage) new key [11$: 0x9002 (S-F1) new key [12$: 0x9003 (S-F2) new key [13$: 0x9004 (S-F3) new key [14$: 0x9005 (S-F4) new key [15$: 0x9006 (S-F5) new key [17$: 0x9007 (S-F6) new key [18$: 0x9008 (S-F7) new key [19$: 0x9009 (S-F8) new key [20$: 0x900a (S-F9) new key [21$: 0x900b (S-F10) new key [23$: 0x900c (S-F11) new key [24$: 0x900d (S-F12) new key [25$: 0x900e (S-F13) new key [26$: 0x900f (S-F14) new key [28$: 0x9010 (S-F15) new key [29$: 0x9011 (S-F16) new key [31$: 0x9012 (S-F17) new key [32$: 0x9013 (S-F18) new key [33$: 0x9014 (S-F19) new key [34$: 0x9015 (S-F20) new key [2$: 0x9016 (S-IC) new key [3$: 0x9017 (S-DC) new key [7$: 0x9018 (S-Home) new key [8$: 0x9019 (S-End) new key [6$: 0x901a (S-NPage) new key [5$: 0x901b (S-PPage) new key [11@: 0xd002 (C-S-F1) new key [12@: 0xd003 (C-S-F2) new key [13@: 0xd004 (C-S-F3) new key [14@: 0xd005 (C-S-F4) new key [15@: 0xd006 (C-S-F5) new key [17@: 0xd007 (C-S-F6) new key [18@: 0xd008 (C-S-F7) new key [19@: 0xd009 (C-S-F8) new key [20@: 0xd00a (C-S-F9) new key [21@: 0xd00b (C-S-F10) new key [23@: 0xd00c (C-S-F11) new key [24@: 0xd00d (C-S-F12) new key [25@: 0xd00e (C-S-F13) new key [26@: 0xd00f (C-S-F14) new key [28@: 0xd010 (C-S-F15) new key [29@: 0xd011 (C-S-F16) new key [31@: 0xd012 (C-S-F17) new key [32@: 0xd013 (C-S-F18) new key [33@: 0xd014 (C-S-F19) new key [34@: 0xd015 (C-S-F20) new key [2@: 0xd016 (C-S-IC) new key [3@: 0xd017 (C-S-DC) new key [7@: 0xd018 (C-S-Home) new key [8@: 0xd019 (C-S-End) new key [6@: 0xd01a (C-S-NPage) new key [5@: 0xd01b (C-S-PPage) new key [I: 0x1031 ((null)) new key [O: 0x1032 ((null)) new key OP: 0x1002 (F1) new key OQ: 0x1003 (F2) new key OR: 0x1004 (F3) new key OS: 0x1005 (F4) new key [15~: 0x1006 (F5) new key [17~: 0x1007 (F6) new key [18~: 0x1008 (F7) new key [19~: 0x1009 (F8) new key [20~: 0x100a (F9) new key [21~: 0x100b (F10) new key [23~: 0x100c (F11) new key [24~: 0x100d (F12) new key [2~: 0x1016 (IC) new key [3~: 0x1017 (DC) replacing key OH: 0x1018 (Home) replacing key OF: 0x1019 (End) new key [6~: 0x101a (NPage) new key [5~: 0x101b (PPage) new key [Z: 0x101c (BTab) replacing key OA: 0x101d (Up) replacing key OB: 0x101e (Down) replacing key OD: 0x101f (Left) replacing key OC: 0x1020 (Right) spawn: /bin/sh -- session 0 created writing 207 to client 7 got 208 from client 7 input_parse: '#' ground input_parse: ' ' ground keys are 7 ([?1;2c) received service class 1 complete key [?1;2c 0xfff keys are 1 (t) complete key t 0x74 input_parse: 't' ground keys are 1 (m) complete key m 0x6d input_parse: 'm' ground keys are 1 (u) complete key u 0x75 input_parse: 'u' ground keys are 1 (x) complete key x 0x78 input_parse: 'x' ground keys are 1 ( ) complete key 0x20 input_parse: ' ' ground keys are 1 (s) complete key s 0x73 input_parse: 's' ground keys are 1 (p) complete key p 0x70 input_parse: 'p' ground keys are 1 (l) complete key l 0x6c input_parse: 'l' ground keys are 1 (i) complete key i 0x69 input_parse: 'i' ground keys are 1 (t) complete key t 0x74 input_parse: 't' ground keys are 1 (-) complete key - 0x2d input_parse: '-' ground keys are 1 (d) complete key d 0x64 input_parse: 'd' ground keys are 1 () complete key 0x7f input_parse: '' ground input_c0_dispatch: ' input_parse: '' ground input_parse: '[' esc_enter input_parse: 'K' csi_enter input_csi_dispatch: 'K' "" "" keys are 1 (w) complete key w 0x77 input_parse: 'w' ground keys are 1 (i) complete key i 0x69 input_parse: 'i' ground keys are 1 (n) complete key n 0x6e input_parse: 'n' ground keys are 1 (d) complete key d 0x64 input_parse: 'd' ground keys are 1 (o) complete key o 0x6f input_parse: 'o' ground keys are 1 (w) complete key w 0x77 input_parse: 'w' ground keys are 1 ( ) complete key 0x20 input_parse: ' ' ground keys are 1 (-) complete key - 0x2d input_parse: '-' ground keys are 1 (h) complete key h 0x68 input_parse: 'h' ground keys are 1 ( ) complete key 0xd input_parse: ' ' ground input_c0_dispatch: ' input_parse: ' ' ground input_c0_dispatch: ' new client 13 got 100 from client 13 got 101 from client 13 got 102 from client 13 got 103 from client 13 got 104 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 105 from client 13 got 106 from client 13 got 200 from client 13 cmdq 0x801c6e160: split-window -h (client 13) spawn: /bin/sh -- writing 203 to client 13 input_parse: '#' ground input_parse: ' ' ground input_parse: '#' ground input_parse: ' ' ground lost client 13 session 0 destroyed writing 203 to client 7 got 205 from client 7 writing 204 to client 7 lost client 7 got 207 from server got 203 from server got 204 from server There are some other peculiarities: With a newly created user (from which I overwrote root's .profile and .shrc, tmux works perfectly. Occasionally (twice out of the 50 or so times I've tested it), the splitting will work fine once in a session. (This happened for example when I ran ktrace on tmux, which I can also post) To explain the 'suddenly' part of the title: when I started my newly updated mysql56-server, tmux immediately exited and lost the session. Recently I changed architectures, from FreeBSD 10.0 i386 to amd64, and I am still working through shared library incompatibilities. I suspect that this could be involved, but I can't imagine how an incompatibility of this sort could result in such a specific, isolated failure.

    Read the article

  • How do client-server cooperation based games like Diablo 3 work?

    - by edgar
    Diablo 3 cooperates with Blizzard servers even during single player games. In fact, Blizzard has had problems with the games "melting their servers." I would like to ask: How do the client and the server communicate? What details does the client leave to the server, and vice versa? What details are redundant - both the client and the server know - and how often do they disagree? The previous paragraph contains the important questions, but I have a few more that I must explain my motivation towards. I am interested in the programming of botting. Ethical botting - I don't plan on actually abusing the automation to run 24/7. I just find it to be a great programming challenge to glean information from a game, and then make decisions from that information. I am stuck in the starting gate. The unofficial questions from this post would be: How can I make a bot (language, tools, libraries)? Can I get information through the communication between client and server, rather than the brute force pixel detection easily used in more static games? There probably is a trust issue, and to that all I can say is that I promise not to abuse the answers. But please feel free to answer any of the questions you feel comfortable with. Thank you!

    Read the article

  • creating ports and initiating communication between client and server

    - by dave
    what i need is a server that listens to 5060 port , when the client sends data to that port the server should open up another port ( any port after 1250 i believe ) and forward the clients data to that port keeping 5060 idle so it can perform the same function for the next client so basically i need the server to a) open up multiple ports one for each client b) get the voice data from the client and be able to send voice data to that client i m looking into the hardware specs and other such details of the scenario so i dont have time to make such a program myself if theres a code that i can run directly ( both server and client side ) on visual studio .net 2010 that will perform these tasks then that would be extremely helpful thanks alot in advance

    Read the article

  • How to implement dual communication between server and client via http

    - by Alex Sebastiano
    I have a AJAX client which must receive messages from server. Some messages from server not like request-response type. For example, imaging game in which players can enter. Server must send to client info about player entering. But how can server send message to client via http without request from client? Only decision that i can invent: client send request to server (getNewPlayerEnter request) with big timeout, server checks state of player set, if in set new players are, then server send info to client, if not server 'sleeps' on some time, and after 'sleeping' server checks players set again. I think my desicion a little stupid(maybe not little). How implement it right? p.s. sorry for my english

    Read the article

  • Wyse Z Class Thin Client (WES7) Image Deployment

    - by nsr81
    We are currently working on deploying a few Wyse WES7 thin clients (Z Class, exact model escapes me at the moment). We have WDM setup, and are able to see new clients show up, interact with them, pull/push images. Our current workflow is something like this: Unpack & boot up the device Disable Windows Firewall, reboot Device shows up in WDM Publish our custom image to the new device. My question is whether or not there is a way I can bypass step 2, and put my image on the client right out of the box. Either using PXE or USB boot disk, such that I can boot into some sort of menu and tell the client to pull a particular image from the WDM server. With HP thin clients, we have a rule setup which detects new clients on the network and pushes our custom image to them first time they show up. I haven't been able to figure it out with Wyse.

    Read the article

  • Not able to connect to a mac client from a windows machine

    - by Manish
    I have a Server.exe file which I use to connect to a mac.(I am fairly confident that server.exe is not buggy ).When i try to do this I get this often cited error "No connection could be made because the target machine actively refused it " I did search some existing questions about this on the forum and it looked like this might be a firewall issue.FWIW I dont have any firewall set on my mac (client) and on my server machine (Windows 7 64 bit ) under the firewall settings I have :- Incoming connections : Block all connections to programs that are not on the list of allowed programs. Active Domain Networks: Same domain as the one which my client is on. Windows Firewire State: Off. Do you think i need to change something here?Can someone help me with next steps?

    Read the article

  • Alternate Client for Cisco Unified Personal Communicator protocol

    - by Jason M
    At work we have an in-house chat system using CUPC. Does anyone else out there use this? There are a few things I do not like about this client: Where's the chat log? If I close the window out, I have no way of getting my conversation back. Tabbed interface? That would be nice. I hate having multiple chat windows up, having to arrange them around my desktop as more people start talking to me. I don't like that I have to use this one-off application for this protocol when other chat clients will handle 99% of the other protocols I use. Tell me: Is the protcol an open standard for which other applications have support? (pidgin, adium, digsby, etc.) If not, can I overcome these issues from within CUPC? Perhaps there are newer versions of the client that overcome these issues.

    Read the article

  • FTP client that supports 2 concurrent FTP sessions

    - by oninea
    I'm looking for an FTP client that can connect to two different FTP servers at the same time and allow file transfer or synchronization between those two servers. Basically what I want to achieve is to transfer/synchronize files between 2 different sites from my local machine. Are there any clients around that support this functionality? If there are none, is there an alternative to achieve this? I've taken a look at net2ftp, a web based FTP client, which provides almost the same functionality that I need. What I'm looking for though is a desktop app. Any ideas?

    Read the article

  • Alternate Client for Cisco Unified Personal Communicator protocol

    - by Jason M
    At work we have an in-house chat system using CUPC. Does anyone else out there use this? There are a few things I do not like about this client: Where's the chat log? If I close the window, I have no way of getting my conversation back. Tabbed interface? That would be nice. I hate having multiple chat windows up, having to arrange them around my desktop as more people start talking to me. I don't like that I have to use this one-off application for particular this protocol when other chat clients will handle 99% of the other protocols I use. Tell me: Is the protocol an open standard for which other applications have support? (pidgin, adium, digsby, etc.) If not, can I overcome these issues from within CUPC? Perhaps there are newer versions of the client that overcome these issues.

    Read the article

  • How can I give my client "full access" to their PHP application's MySQL database?

    - by Micah Delane Bolen
    I am building a PHP application for a client and I'm seriously considering WordPress or a simple framework that will allow me to quickly build out features like forums, etc. However, the client is adamant about having "full access" to the database and the ability to "mine the data." Unfortunately, I'm almost certain they will be disappointed when they realize they won't be able to easily glean meaningful insight by looking at serialized fields in wp_usermeta, etc. One thought I had was to replicate a variation on the live database where I flatten out all of those ambiguous and/or serialized fields into something that is then parsable by a mere mortal using a tool as simple as phpMyAdmin. Unfortunately, the client is not going to settle for a simple backend dashboard where I create the custom reports for them even though I know that would be the easiest and most sane approach.

    Read the article

  • Benefits of using the same language for client and server?

    - by Makita
    I'm looking at architecture solutions for a mobile project that will have a web-service/app in addition to native apps. I've been looking at various libraries, frameworks, and stacks like jqm, backbone, parse, and meteor. Meteor, sort of an "open stack package framework", is tightly bound with node.js. There is a lot of talk about the benefits of using the same language both client and server side, and I'm not getting it. I could understand if you want to mirror the entire state of a web application on both client and server but struggling to find other wins... Workflow efficiency? I'm trying to understand why client/server language parity is considered to be a holy grail, any explicit examples or links would be greatly appreciated, thanks!

    Read the article

  • OpenVPN - client-to-client traffic working in one direction but not the other

    - by user42055
    I have the following VPN configuration: +------------+ +------------+ +------------+ | outpost |----------------| kino |----------------| guchuko | +------------+ +------------+ +------------+ OS: FreeBSD 6.2 OS: Gentoo 2.6.32 OS: Gentoo 2.6.33.3 Keyname: client3 Keyname: server Keyname: client1 eth0: 10.0.1.254 eth0: 203.x.x.x eth0: 192.168.0.6 tun0: 192.168.150.18 tun0: 192.168.150.1 tun0: 192.168.150.10 P-t-P: 192.166.150.17 P-t-P: 192.168.150.2 P-t-P: 192.168.150.9 Kino is the server and has client-to-client enabled. All three machines have ip forwarding enabled, by this on the gentoo boxes: net.ipv4.conf.all.forwarding = 1 And this on the FreeBSD box: net.inet.ip.forwarding: 1 In the server's "ccd" directory is the following files: client1: iroute 192.168.0.0 255.255.255.0 client3: iroute 10.0.1.0 255.255.255.0 The server config has these routes configured: push "route 192.168.0.0 255.255.255.0" push "route 10.0.1.0 255.255.255.0" route 192.168.0.0 255.255.255.0 route 10.0.1.0 255.255.255.0 Kino's routing table looks like this: 192.168.150.0 192.168.150.2 255.255.255.0 UG 0 0 0 tun0 10.0.1.0 192.168.150.2 255.255.255.0 UG 0 0 0 tun0 192.168.0.0 192.168.150.2 255.255.255.0 UG 0 0 0 tun0 192.168.150.2 0.0.0.0 255.255.255.255 UH 0 0 0 tun0 Outpost's like this: 192.168.150 192.168.150.17 UGS 0 17 tun0 192.168.0 192.168.150.17 UGS 0 2 tun0 192.168.150.17 192.168.150.18 UH 3 0 tun0 And Guchuko's like this: 192.168.150.0 192.168.150.9 255.255.255.0 UG 0 0 0 tun0 10.0.1.0 192.168.150.9 255.255.255.0 UG 0 0 0 tun0 192.168.150.9 0.0.0.0 255.255.255.255 UH 0 0 0 tun0 Now, the tests. Pings from Guchuko to Outpost's LAN IP work OK, as does the reverse - pings from Outpost to Guchuko's LAN IP. However... Pings from Outpost, to a machine on Guchuko's LAN work fine: .(( root@outpost )). (( 06:39 PM )) :: ~ :: # ping 192.168.0.3 PING 192.168.0.3 (192.168.0.3): 56 data bytes 64 bytes from 192.168.0.3: icmp_seq=0 ttl=63 time=462.641 ms 64 bytes from 192.168.0.3: icmp_seq=1 ttl=63 time=557.909 ms But a ping from Guchuko, to a machine on Outpost's LAN does not: .(( root@guchuko )). (( 06:43 PM )) :: ~ :: # ping 10.0.1.253 PING 10.0.1.253 (10.0.1.253) 56(84) bytes of data. --- 10.0.1.253 ping statistics --- 3 packets transmitted, 0 received, 100% packet loss, time 2000ms Guchuko's tcpdump of tun0 shows: 18:46:27.716931 IP 192.168.150.10 > 10.0.1.253: ICMP echo request, id 63009, seq 1, length 64 18:46:28.716715 IP 192.168.150.10 > 10.0.1.253: ICMP echo request, id 63009, seq 2, length 64 18:46:29.716714 IP 192.168.150.10 > 10.0.1.253: ICMP echo request, id 63009, seq 3, length 64 Outpost's tcpdump on tun0 shows: 18:44:00.333341 IP 192.168.150.10 > 10.0.1.253: ICMP echo request, id 63009, seq 3, length 64 18:44:01.334073 IP 192.168.150.10 > 10.0.1.253: ICMP echo request, id 63009, seq 4, length 64 18:44:02.331849 IP 192.168.150.10 > 10.0.1.253: ICMP echo request, id 63009, seq 5, length 64 So Outpost is receiving the ICMP request destined for the machine on it's subnet, but appears not be forwarding it. Outpost has gateway_enable="YES" in its rc.conf which correctly sets net.inet.ip.forwarding to 1 as mentioned earlier. As far as I know, that's all that's required to make a FreeBSD box forward packets between interfaces. Is there something else I could be forgetting ?

    Read the article

  • Eventlet or gevent or Stackless + Twisted, Pylons, Django and SQL Alchemy

    - by Khorkrak
    We're using Twisted extensively for apps requiring a great deal of asynchronous io. There are some cases where stuff is cpu bound instead and for that we spawn a pool of processes to do the work and have a system for managing these across multiple servers as well - all done in Twisted. Works great. The problem is that it's hard to bring new team members up to speed. Writing asynchronous code in Twisted requires a near vertical learning curve. It's as if humans just don't think that way naturally. We're considering a mixed approach perhaps. Maybe do the xmlrpc server part and process management in Twisted still but the other stuff in code that at least looks synchronous while not being as such. Then again I like explicit over implicit so hmmm. Anyway onto greenlets - how well does that stuff work? So there's Stackless and as you can see from my Gallentean avatar I'm well aware of the tremendous success in it's use for CCP's flagship EVE Online game first hand. What about Eventlet or gevent? Well for now only Eventlet works with Twisted. However gevent claims to be faster since it's not a pure python implementation it instead uses libevent. It also has fewer idiosyncrasies and defects supposedly. The documentation there is minimal in comparison to Eventlet and it's maintained by 1 guy as far as I can tell. This makes me leery but all great projects start this way so... Then there's PyPy - I haven't even finished reading about that one yet - just saw it in this thread: Drawbacks of Stackless. So confusing - I'm wondering what the heck to do - sounds like Eventlet is probably the best bet but is it really stable enough? Anyone out there have any experience with it? Should we go with Stackless instead as it's been around and is proven technology - just like Twisted is as well - and they do work together nicely. But still I hate having to have a separate version of Python to do this. what to do.... This somewhat obnoxious blog entry hit the nail on the head for me though: Asynchronous IO for Grownups We're stuck using MySQL as well - I never knew how great PostgreSQL was until having had to work on a production OLTP system in MySQL instead - but that's another story. But if that monkey patch thing really works then wow. Just wow.

    Read the article

  • Caveats of select/poll vs. epoll reactors in Twisted

    - by David
    Everything I've read and experienced ( Tornado based apps ) leads me to believe that ePoll is a natural replacement for Select and Poll based networking, especially with Twisted. Which makes me paranoid, its pretty rare for a better technique or methodology not to come with a price. Reading a couple dozen comparisons between epoll and alternatives shows that epoll is clearly the champion for speed and scalability, specifically that it scales in a linear fashion which is fantastic. That said, what about processor and memory utilization, is epoll still the champ?

    Read the article

  • Twisted Python getPage

    - by David Dixon II
    I tried to get support on this but I am TOTALLY confused. Here's my code: from twisted.internet import reactor from twisted.web.client import getPage from twisted.web.error import Error from twisted.internet.defer import DeferredList from sys import argv class GrabPage: def __init__(self, page): self.page = page def start(self, *args): if args == (): # We apparently don't need authentication for this d1 = getPage(self.page) else: if len(args) == 2: # We have our login information d1 = getPage(self.page, headers={"Authorization": " ".join(args)}) else: raise Exception('Missing parameters') d1.addCallback(self.pageCallback) dl = DeferredList([d1]) d1.addErrback(self.errorHandler) dl.addCallback(self.listCallback) def errorHandler(self,result): # Bad thingy! pass def pageCallback(self, result): return result def listCallback(self, result): print result a = GrabPage('http://www.google.com') data = a.start() # Not the HTML I wish to get the HTML out which is given to pageCallback when start() is called. This has been a pita for me. Ty! And sorry for my sucky coding.

    Read the article

  • Plesk 9 - Client panel access via client domain, not server host name

    - by Ben
    I've setup a Plesk 9 vps recently. The plesk admin is accessible at, say, http://superawesomedomain.com:8443. I'm setting up a couple of client domains on the server. I'd like clients to be able to access plesk via a similar url, but instead it would be something like http://inferiorclientdomain.com:8443. This works ok in a sense, but it always redirect the request for inferiorclientdomain.com:8443 to superawesomedomain.com:8443. I'd like the control panel access url to remain on inferiorclientdomain.com. Is that possible? EDIT: System Info. Plesk 9.5.2 on CentOS 5 (Build 95100504.12)

    Read the article

  • Clearos open vpn vs windows open vpn client where client connects with no default gateway

    - by Paul
    Am using clearos as open vpn server and configured my users on windows machine with open vpn client. My problem is that users connect to the server without a default gateway and also with ip conflicts, i can ping the server but i can not ping any user behind the server. please any one can help to find out what causes the clients to connect without a default gateway and also not to be able to ping any user behind the clearos open vpn server. Help with a step by step guide of installing open vpn on clearos and open vpn clients on windows. Thanks

    Read the article

  • OpenVPN - client-to-client traffic working in one direction but not the other

    - by Pawz
    I have the following VPN configuration: +------------+ +------------+ +------------+ | outpost |----------------| kino |----------------| guchuko | +------------+ +------------+ +------------+ OS: FreeBSD 6.2 OS: Gentoo 2.6.32 OS: Gentoo 2.6.33.3 Keyname: client3 Keyname: server Keyname: client1 eth0: 10.0.1.254 eth0: 203.x.x.x eth0: 192.168.0.6 tun0: 192.168.150.18 tun0: 192.168.150.1 tun0: 192.168.150.10 P-t-P: 192.166.150.17 P-t-P: 192.168.150.2 P-t-P: 192.168.150.9 Kino is the server and has client-to-client enabled. I am using "fragment 1400" and "mssfix" on all three machines. An mtu-test on both connections is successful. All three machines have ip forwarding enabled, by this on the gentoo boxes: net.ipv4.conf.all.forwarding = 1 And this on the FreeBSD box: net.inet.ip.forwarding: 1 In the server's "ccd" directory is the following files: client1: iroute 192.168.0.0 255.255.255.0 client3: iroute 10.0.1.0 255.255.255.0 The server config has these routes configured: push "route 192.168.0.0 255.255.255.0" push "route 10.0.1.0 255.255.255.0" route 192.168.0.0 255.255.255.0 route 10.0.1.0 255.255.255.0 Kino's routing table looks like this: 192.168.150.0 192.168.150.2 255.255.255.0 UG 0 0 0 tun0 10.0.1.0 192.168.150.2 255.255.255.0 UG 0 0 0 tun0 192.168.0.0 192.168.150.2 255.255.255.0 UG 0 0 0 tun0 192.168.150.2 0.0.0.0 255.255.255.255 UH 0 0 0 tun0 Outpost's like this: 192.168.150 192.168.150.17 UGS 0 17 tun0 192.168.0 192.168.150.17 UGS 0 2 tun0 192.168.150.17 192.168.150.18 UH 3 0 tun0 And Guchuko's like this: 192.168.150.0 192.168.150.9 255.255.255.0 UG 0 0 0 tun0 10.0.1.0 192.168.150.9 255.255.255.0 UG 0 0 0 tun0 192.168.150.9 0.0.0.0 255.255.255.255 UH 0 0 0 tun0 Now, the tests. Pings from Guchuko to Outpost's LAN IP work OK, as does the reverse - pings from Outpost to Guchuko's LAN IP. However... Pings from Outpost, to a machine on Guchuko's LAN work fine: .(( root@outpost )). (( 06:39 PM )) :: ~ :: # ping 192.168.0.3 PING 192.168.0.3 (192.168.0.3): 56 data bytes 64 bytes from 192.168.0.3: icmp_seq=0 ttl=63 time=462.641 ms 64 bytes from 192.168.0.3: icmp_seq=1 ttl=63 time=557.909 ms But a ping from Guchuko, to a machine on Outpost's LAN does not: .(( root@guchuko )). (( 06:43 PM )) :: ~ :: # ping 10.0.1.253 PING 10.0.1.253 (10.0.1.253) 56(84) bytes of data. --- 10.0.1.253 ping statistics --- 3 packets transmitted, 0 received, 100% packet loss, time 2000ms Guchuko's tcpdump of tun0 shows: 18:46:27.716931 IP 192.168.150.10 > 10.0.1.253: ICMP echo request, id 63009, seq 1, length 64 18:46:28.716715 IP 192.168.150.10 > 10.0.1.253: ICMP echo request, id 63009, seq 2, length 64 18:46:29.716714 IP 192.168.150.10 > 10.0.1.253: ICMP echo request, id 63009, seq 3, length 64 Outpost's tcpdump on tun0 shows: 18:44:00.333341 IP 192.168.150.10 > 10.0.1.253: ICMP echo request, id 63009, seq 3, length 64 18:44:01.334073 IP 192.168.150.10 > 10.0.1.253: ICMP echo request, id 63009, seq 4, length 64 18:44:02.331849 IP 192.168.150.10 > 10.0.1.253: ICMP echo request, id 63009, seq 5, length 64 So Outpost is receiving the ICMP request destined for the machine on it's subnet, but appears not be forwarding it. Outpost has gateway_enable="YES" in its rc.conf which correctly sets net.inet.ip.forwarding to 1 as mentioned earlier. As far as I know, that's all that's required to make a FreeBSD box forward packets between interfaces. Is there something else I could be forgetting ? FWIW, pinging 10.0.1.253 from Kino has the same result - the traffic does not get forwarded. UPDATE: I've found that I can only ping certain IP's on Guchuko's LAN from Outpost. From Outpost I can ping 192.168.0.3 and 192.168.0.2, but 192.168.99 and 192.168.0.4 are unreachable. The same tcpdump behavior can be seen. I think this means the problem can't be due to ipforwarding or routing, because Outpost can reach SOME hosts on Guchuko's LAN but not others and likewise, Guchuko can reach two hosts on Outpost's LAN, but not others. This baffles me.

    Read the article

  • Cisco AnyConnect VPN client - prevent connecting as work network

    - by Opmet
    From Windows 7 I'm using "Cisco AnyConnect Secure Mobility Client 3.0" to connect to our corporate network. Every time I establish the VPN connection Windows will set the type as "work network". I don't want this. So I go to "network and sharing center" and manually / interactively change it to "public network". But I have to repeat it for every new VPN connection. Is there any way to make Windows remember / persist this configuration? Can it be configured in the VPN client? Do our IT admins need to change something at server end? Motivation: A "work network" per default uses different firewall settings that allows for stuff like "network discovery" and "file shares". But I just need "remote desktop" (mstsc). Additional info: Our IT admins claimed this would be Windows default behaviour and there was nothing we could do about it: Windows would always initiate a VPN connection as "work network". Based on this statement I assume this is a "general" issue and went ahead posting here (at superuser.com).

    Read the article

  • CentOS 5.4 NFS v4 client file permissions differ from original files & NFS Share file contents

    - by p4guru
    Having a strange problem with NFS share and file permissions on the 1 out of the 2 NFS clients, web1 has file permissions issues but web2 is fine. web1 and web2 are load balanced web servers. So questions are: how do I ensure NFS share file contents retain the same permissions for user/group as the original files on web1 server like they do on web2 server ? how do I reverse what I did on web1, i tried unmount command and said command not found ? Information: I'm using 3 dedicated server setup. All 3 servers CentOS 5.4 64bit based. servers are as follows: web1 - nfs client with file permissions issues web2 - nfs client file permissions are OKAY db1 - nfs share at /nfsroot web2 nfs client was setup by my web host, while web1 was setup by me. I did the following commands on web1 and it worked with updating db1 nfsroot share at /nfsroot/site_css with latest files on web1 but the file permissions don't stick even if i use tar with -p command to perserve file permissions ? cd /home/username/public_html/forums/script/ tar -zcp site_css/ > site_css.tar.gz mount -t nfs4 nfsshareipaddress:/site_css /home/username/public_html/forums/scripts/site_css/ -o rw,soft cd /home/username/public_html/forums/script/ tar -zxf site_css.tar.gz But checking on web1 file permissions no longer username user/group but owned by nobody ? but web2 file permissions correct ? This is only a problem for web1 while web2 is correct ? Looks like numeric ids aren't the same ? Not sure how to correct this ? web1 with incorrect user/group of nobody ls -alh /home/username/public_html/forums/scripts/site_css total 48K drwxrwxrwx 2 nobody nobody 4.0K Feb 22 02:37 ./ drwxr-xr-x 3 username username 4.0K Feb 22 02:43 ../ -rw-r--r-- 1 nobody nobody 1 Nov 30 2006 index.html -rw-r--r-- 1 nobody nobody 5.8K Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 nobody nobody 5.8K Feb 22 02:37 style-95001864-00002.css -rw-r--r-- 1 nobody nobody 5.8K Feb 18 05:37 style-b1879ba7-00002.css -rw-r--r-- 1 nobody nobody 5.8K Feb 18 05:37 style-cc2f96c9-00011.css web1 numeric ids ls -n /home/username/public_html/forums/scripts/site_css total 48 drwxrwxrwx 2 99 99 4096 Feb 22 02:37 ./ drwxr-xr-x 3 503 500 4096 Feb 22 02:43 ../ -rw-r--r-- 1 99 99 1 Nov 30 2006 index.html -rw-r--r-- 1 99 99 5876 Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 99 99 5877 Feb 22 02:37 style-95001864-00002.css -rw-r--r-- 1 99 99 5877 Feb 18 05:37 style-b1879ba7-00002.css -rw-r--r-- 1 99 99 5876 Feb 18 05:37 style-cc2f96c9-00011.css web2 correct username user/group permissions ls -alh /home/username/public_html/forums/scripts/site_css total 48K drwxrwxrwx 2 root root 4.0K Feb 22 02:37 ./ drwxr-xr-x 3 username username 4.0K Dec 2 14:51 ../ -rw-r--r-- 1 username username 1 Nov 30 2006 index.html -rw-r--r-- 1 username username 5.8K Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 username username 5.8K Feb 22 02:37 style-95001864-00002.css -rw-r--r-- 1 username username 5.8K Feb 18 05:37 style-b1879ba7-00002.css -rw-r--r-- 1 username username 5.8K Feb 18 05:37 style-cc2f96c9-00011.css web2 numeric ids ls -n /home/username/public_html/forums/scripts/site_css total 48 drwxrwxrwx 2 503 500 4096 Feb 22 02:37 ./ drwxr-xr-x 3 503 500 4096 Dec 2 14:51 ../ -rw-r--r-- 1 503 500 1 Nov 30 2006 index.html -rw-r--r-- 1 503 500 5876 Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 503 500 5877 Feb 22 02:37 style-95001864-00002.css -rw-r--r-- 1 503 500 5877 Feb 18 05:37 style-b1879ba7-00002.css -rw-r--r-- 1 503 500 5876 Feb 18 05:37 style-cc2f96c9-00011.css I checked db1 /nfsroot/site_css and user/group ownership was incorrect for newer files dated feb22 owned by root and not username ? on db1 originally incorrect root assigned user/group for new feb22 dated files ls -alh /nfsroot/site_css total 44K drwxrwxrwx 2 root root 4.0K Feb 22 02:37 . drwxr-xr-x 17 root root 4.0K Feb 17 12:06 .. -rw-r--r-- 1 root root 1 Nov 30 2006 index.html -rw-r--r-- 1 root root 5.8K Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 root root 5.8K Feb 22 02:37 style-95001864-00002.css -rw------- 1 username nfs 5.8K Feb 18 05:37 style-b1879ba7-00002.css -rw------- 1 username nfs 5.8K Feb 18 05:37 style-cc2f96c9-00011.css Then I chmod them all on db1 and chown to set to right ownership on db1 so it looks like below on db1 once corrected the newer feb22 dated files ls -alh /nfsroot/site_css total 44K drwxrwxrwx 2 root root 4.0K Feb 22 02:37 . drwxr-xr-x 17 root root 4.0K Feb 17 12:06 .. -rw-r--r-- 1 username username 1 Nov 30 2006 index.html -rw-r--r-- 1 username username 5.8K Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 username username 5.8K Feb 22 02:37 style-95001864-00002.css -rw-r--r-- 1 username username 5.8K Feb 18 05:37 style-b1879ba7-00002.css -rw-r--r-- 1 username username 5.8K Feb 18 05:37 style-cc2f96c9-00011.css but still web1 shows owned by nobody ? while web2 shows correct permissions ? web1 still with incorrect user/group of nobody not matching what web2 and db1 are set to ? ls -alh /home/username/public_html/forums/scripts/site_css total 48K drwxrwxrwx 2 nobody nobody 4.0K Feb 22 02:37 ./ drwxr-xr-x 3 username username 4.0K Feb 22 02:43 ../ -rw-r--r-- 1 nobody nobody 1 Nov 30 2006 index.html -rw-r--r-- 1 nobody nobody 5.8K Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 nobody nobody 5.8K Feb 22 02:37 style-95001864-00002.css -rw-r--r-- 1 nobody nobody 5.8K Feb 18 05:37 style-b1879ba7-00002.css -rw-r--r-- 1 nobody nobody 5.8K Feb 18 05:37 style-cc2f96c9-00011.css Just so confusing so any help is very very much appreciated! thanks

    Read the article

  • CentOS 5.4 NFS v4 client file permissions differ from original files & NFS Share file contents

    - by p4guru
    Having a strange problem with NFS share and file permissions on the 1 out of the 2 NFS clients, web1 has file permissions issues but web2 is fine. web1 and web2 are load balanced web servers. So questions are: how do I ensure NFS share file contents retain the same permissions for user/group as the original files on web1 server like they do on web2 server ? how do I reverse what I did on web1, i tried unmount command and said command not found ? Information: I'm using 3 dedicated server setup. All 3 servers CentOS 5.4 64bit based. servers are as follows: web1 - nfs client with file permissions issues web2 - nfs client file permissions are OKAY db1 - nfs share at /nfsroot web2 nfs client was setup by my web host, while web1 was setup by me. I did the following commands on web1 and it worked with updating db1 nfsroot share at /nfsroot/site_css with latest files on web1 but the file permissions don't stick even if i use tar with -p command to perserve file permissions ? cd /home/username/public_html/forums/script/ tar -zcp site_css/ > site_css.tar.gz mount -t nfs4 nfsshareipaddress:/site_css /home/username/public_html/forums/scripts/site_css/ -o rw,soft cd /home/username/public_html/forums/script/ tar -zxf site_css.tar.gz But checking on web1 file permissions no longer username user/group but owned by nobody ? but web2 file permissions correct ? This is only a problem for web1 while web2 is correct ? Looks like numeric ids aren't the same ? Not sure how to correct this ? web1 with incorrect user/group of nobody ls -alh /home/username/public_html/forums/scripts/site_css total 48K drwxrwxrwx 2 nobody nobody 4.0K Feb 22 02:37 ./ drwxr-xr-x 3 username username 4.0K Feb 22 02:43 ../ -rw-r--r-- 1 nobody nobody 1 Nov 30 2006 index.html -rw-r--r-- 1 nobody nobody 5.8K Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 nobody nobody 5.8K Feb 22 02:37 style-95001864-00002.css -rw-r--r-- 1 nobody nobody 5.8K Feb 18 05:37 style-b1879ba7-00002.css -rw-r--r-- 1 nobody nobody 5.8K Feb 18 05:37 style-cc2f96c9-00011.css web1 numeric ids ls -n /home/username/public_html/forums/scripts/site_css total 48 drwxrwxrwx 2 99 99 4096 Feb 22 02:37 ./ drwxr-xr-x 3 503 500 4096 Feb 22 02:43 ../ -rw-r--r-- 1 99 99 1 Nov 30 2006 index.html -rw-r--r-- 1 99 99 5876 Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 99 99 5877 Feb 22 02:37 style-95001864-00002.css -rw-r--r-- 1 99 99 5877 Feb 18 05:37 style-b1879ba7-00002.css -rw-r--r-- 1 99 99 5876 Feb 18 05:37 style-cc2f96c9-00011.css web2 correct username user/group permissions ls -alh /home/username/public_html/forums/scripts/site_css total 48K drwxrwxrwx 2 root root 4.0K Feb 22 02:37 ./ drwxr-xr-x 3 username username 4.0K Dec 2 14:51 ../ -rw-r--r-- 1 username username 1 Nov 30 2006 index.html -rw-r--r-- 1 username username 5.8K Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 username username 5.8K Feb 22 02:37 style-95001864-00002.css -rw-r--r-- 1 username username 5.8K Feb 18 05:37 style-b1879ba7-00002.css -rw-r--r-- 1 username username 5.8K Feb 18 05:37 style-cc2f96c9-00011.css web2 numeric ids ls -n /home/username/public_html/forums/scripts/site_css total 48 drwxrwxrwx 2 503 500 4096 Feb 22 02:37 ./ drwxr-xr-x 3 503 500 4096 Dec 2 14:51 ../ -rw-r--r-- 1 503 500 1 Nov 30 2006 index.html -rw-r--r-- 1 503 500 5876 Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 503 500 5877 Feb 22 02:37 style-95001864-00002.css -rw-r--r-- 1 503 500 5877 Feb 18 05:37 style-b1879ba7-00002.css -rw-r--r-- 1 503 500 5876 Feb 18 05:37 style-cc2f96c9-00011.css I checked db1 /nfsroot/site_css and user/group ownership was incorrect for newer files dated feb22 owned by root and not username ? on db1 originally incorrect root assigned user/group for new feb22 dated files ls -alh /nfsroot/site_css total 44K drwxrwxrwx 2 root root 4.0K Feb 22 02:37 . drwxr-xr-x 17 root root 4.0K Feb 17 12:06 .. -rw-r--r-- 1 root root 1 Nov 30 2006 index.html -rw-r--r-- 1 root root 5.8K Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 root root 5.8K Feb 22 02:37 style-95001864-00002.css -rw------- 1 username nfs 5.8K Feb 18 05:37 style-b1879ba7-00002.css -rw------- 1 username nfs 5.8K Feb 18 05:37 style-cc2f96c9-00011.css Then I chmod them all on db1 and chown to set to right ownership on db1 so it looks like below on db1 once corrected the newer feb22 dated files ls -alh /nfsroot/site_css total 44K drwxrwxrwx 2 root root 4.0K Feb 22 02:37 . drwxr-xr-x 17 root root 4.0K Feb 17 12:06 .. -rw-r--r-- 1 username username 1 Nov 30 2006 index.html -rw-r--r-- 1 username username 5.8K Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 username username 5.8K Feb 22 02:37 style-95001864-00002.css -rw-r--r-- 1 username username 5.8K Feb 18 05:37 style-b1879ba7-00002.css -rw-r--r-- 1 username username 5.8K Feb 18 05:37 style-cc2f96c9-00011.css but still web1 shows owned by nobody ? while web2 shows correct permissions ? web1 still with incorrect user/group of nobody not matching what web2 and db1 are set to ? ls -alh /home/username/public_html/forums/scripts/site_css total 48K drwxrwxrwx 2 nobody nobody 4.0K Feb 22 02:37 ./ drwxr-xr-x 3 username username 4.0K Feb 22 02:43 ../ -rw-r--r-- 1 nobody nobody 1 Nov 30 2006 index.html -rw-r--r-- 1 nobody nobody 5.8K Feb 22 02:37 style-057c3df0-00011.css -rw-r--r-- 1 nobody nobody 5.8K Feb 22 02:37 style-95001864-00002.css -rw-r--r-- 1 nobody nobody 5.8K Feb 18 05:37 style-b1879ba7-00002.css -rw-r--r-- 1 nobody nobody 5.8K Feb 18 05:37 style-cc2f96c9-00011.css Just so confusing so any help is very very much appreciated! thanks

    Read the article

< Previous Page | 3 4 5 6 7 8 9 10 11 12 13 14  | Next Page >