Search Results

Search found 10941 results on 438 pages for 'location'.

Page 103/438 | < Previous Page | 99 100 101 102 103 104 105 106 107 108 109 110  | Next Page >

  • Odd squid transparent redirect behavior

    - by EMiller
    This is the first time I've set up squid. It's running a redirect script that does some text search/replace on html pages, and then saves them to a location on the same machine on the nginx path - then issues the redirect to that URL (it's an art project :D). The relevant lines in squid.conf are http_port 3128 transparent redirect_program /etc/squid/jefferson_redirect.py The jefferson_redirect.py script is based on this script: http://gofedora.com/how-to-write-custom-redirector-rewritor-plugin-squid-python/ The issue: I'm getting strange http redirect behavior. For example, here is the normal request/response from a PHP script that issues a header("Location:"); - a 302 redirect: http://redirector.mysite.com/?unicmd=g+yreka GET /?unicmd=g+yreka HTTP/1.1 Host: redirector.mysite.com User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.1.9) Gecko/20100330 Fedora/3.5.9-1.fc12 Firefox/3.5.9 Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8 Accept-Language: en-us,en;q=0.5 Accept-Encoding: gzip,deflate Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7 Keep-Alive: 300 Connection: keep-alive HTTP/1.1 302 Found Date: Tue, 13 Apr 2010 05:15:43 GMT Server: Apache X-Powered-By: PHP/5.2.11 Location: http://www.google.com/search?q=yreka Content-Type: text/html Vary: User-Agent,Accept-Encoding Content-Encoding: gzip Content-Length: 2108 Keep-Alive: timeout=3, max=100 Connection: Keep-Alive Here's what it looks like when running through the squid proxy (note that "redirector.mysite.com" is not the site running squid or nginx): http://redirector.mysite.com/?unicmd=g+yreka GET /?unicmd=g+yreka HTTP/1.1 Host: redirector.mysite.com User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.1.9) Gecko/20100330 Fedora/3.5.9-1.fc12 Firefox/3.5.9 Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8 Accept-Language: en-us,en;q=0.5 Accept-Encoding: gzip,deflate Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7 Keep-Alive: 300 Proxy-Connection: keep-alive If-Modified-Since: Tue, 13 Apr 2010 05:21:02 GMT HTTP/1.0 200 OK Server: nginx/0.7.62 Date: Tue, 13 Apr 2010 05:21:10 GMT Content-Type: text/html Content-Length: 17865 Last-Modified: Tue, 13 Apr 2010 05:21:10 GMT Accept-Ranges: bytes X-Cache: MISS from jefferson X-Cache-Lookup: HIT from jefferson:3128 Via: 1.1 jefferson:3128 (squid/2.7.STABLE6) Connection: keep-alive Proxy-Connection: keep-alive It is basically working - but the URL http://redirector.mysite.com/?unicmd=g+yreka remains unchanged, while displaying the google page (mostly broken as it's using URLs relative to redirector.mysite.com) I've experienced a similar thing with google results pages: when clicking to another page from google, I get a google URL, with the other site's content. Sorry for the long post - many thanks if you've read this far! Any ideas?

    Read the article

  • Subversion problem, repo has moved

    - by Rudiger
    Hi, I've set up subversion on a CentOS fresh install. Web view works fine and gives no errors and requests password but when I try and access it through svn client (xcode) it gives the error 175011 (Repository has been moved). I've tried some of the solutions out there but no success. My subversion.conf: <Location /repos> DAV svn SVNParentPath /var/www/html/repos # Limit write permission to list of valid users. # Require SSL connection for password protection. SSLRequireSSL AuthType Basic AuthName "Authorization Realm" AuthUserFile /etc/svn-auth-conf Require valid-user </Location> My Apache DocumentRoot: /var/www/html I've only set up one svn repository so far so there shouldn't be any conflicts there. If you need any more info let me know. Thanks

    Read the article

  • Connecting a network drive only when the VPN is connected?

    - by leeand00
    I have a thumb drive that I want to be able to connect to the same place over the network at one location as it is locally. Sometimes I leave it at the other location, but usually if I'm going to back it up, I back it up locally for network traffic. Is there a way to automatically connect the thumb drive when I connect to the VPN? (Always to the same drive letter...and obviously skip connecting it if it's already plugged in locally and the VPN is connected...) I'm using a Cisco VPN Client 5.x

    Read the article

  • Nginx static files exclude one or some file extensions

    - by Evgeniy
    I'm serving up a static site via nginx. location ~* \.(avi|bin|bmp|dmg|doc|docx|dpkg|exe|flv|gif|htm|html|ico|ics|img|jpeg|jpg|m2a|m2v|mov|mp3|mp4|mpeg|mpg|msi|pdf|pkg|png|ppt|pptx|ps|rar|rss|rtf|swf|tif|tiff|txt|wmv|xhtml|xls|xml|zip)$ { root /var/www/html1; access_log off; expires 1d; } And my goal is to exclude requests like http://connect1.webinar.ru/converter/task/. Full view is like http://mydomain.tld/converter/task/setComplete/fid/34330/fn/7c2cfed32ec2eef6788e728fa46f7a80.ppt.swf. Despite the fact these URLs ends in such a format they are not static, but fake script requests, so I have a problems with them. What is the best way to do this? How can I add an exclusion for this URL or maybe I can to exclude the specific file exptension (.ppt.swf, pptx.swf) from the list of this Nginx location? Thanks.

    Read the article

  • Running .net application over a network

    - by Marlon
    Hello, I need some advice please. I need to enable a .Net application to run over a network share, the problem is that this will be on clients network shares and so the path will not be identical. I've had a quick look at ClickOnce and the publish options in VS2008 but it wants a specific network share location - and I'm assuming this location gets stored somewhere when it does its thing. At the moment the job is being done with a old VB6 application and so gets around all these security issues, but that application is poorly written and almost impossible to maintain so it really needs to go. Is it possible for the domain controller to be set up to allow this specific .Net application to execute? Any other options would be welcomed as I want to get this little application is very business critical. I aught to say that the client networks are schools, and thus are often quite locked down as are the client machines, so manually adding exceptions to each client machine is a big no no. Marlon

    Read the article

  • How to disable proxy cache when query string is empty?

    - by chx
    With nginx I have server { listen 1.2.3.4:80 proxy_cache_valid 200 302 5m; location / { try_files $uri @upstream; root $root; } } When I go http://example.com/foobar it generates a redirect to http://example.com/foobar?filter_distance=50&... which is visitor dependent so I would like to not cache this redirect. I need to bypass cache when the query string is empty. I am a bit lost because location /foobar will match both.

    Read the article

  • In Nginx can I set Keep-Alive dynamically depending on ssl connection?

    - by ck_
    I would like to avoid having to repeat all the virtualhost server {} blocks in nginx just to have custom ssl settings that vary slightly from plain http requests. Most ssl directives can be placed right in the main block, except one hurdle I cannot find a workaround for: different keep-alive for https vs http Is there any way I can use $scheme to dynamically change the keepalive_timeout ? I've even considered that I can use more_set_input_headers -r 'Keep-Alive: timeout=60'; to conditionally replace the keep-alive timeout only if it already exists, but the problem is $scheme cannot be used in location ie. this is invalid location ^https {}

    Read the article

  • Changing Network Path of Offline Files

    - by Adam
    Many of our users have their Home folder set as Available Offline. Their Windows 7 laptops will not be back on our network for a few weeks. In the mean time, we're setting up new servers and reorganizing our files, so the network path to the Home folder is going to be completely different. Based on some testing I did, when the users return, any files they've created or modified while offline will be gone, and the new Home folder will be there and not set to sync. The offline cache of the old Home folder is still accessible through the Sync Center, but they're not going to want to dig through that and try to find what's missing. Avoiding this would involve keeping the old server around and moving everyone to the new location in person, so we know for sure they're synced first. Is there any way to avoid this that isn't as tedious, like a quick registry edit or something that will point the old offline cache to the new location?

    Read the article

  • nginx crashes on ssl after about a minute

    - by Scott
    Here are my configuration files ssl.conf # HTTPS server # server { listen 443 ssl; server_name api.domain.com; error_log /var/log/nginx/api.error.log; location / { root /var/www/api.domain.com; index index.html index.php index.php; try_files $uri $uri/ /index.php?$args; } ssl on; ssl_certificate /etc/nginx/api.domain.com.crt; ssl_certificate_key /etc/nginx/api.domain.com.key; ssl_session_timeout 5m; ssl_protocols SSLv2 SSLv3 TLSv1; ssl_ciphers HIGH:!aNULL:!MD5; ssl_prefer_server_ciphers on; # pass the PHP scripts to FastCGI server listening on 127.0.0.1:9000 location ~ \.php$ { # root html; fastcgi_pass 127.0.0.1:9000; fastcgi_index index.php; fastcgi_split_path_info ^(.+\.php)(.*)$; fastcgi_param SCRIPT_FILENAME /var/www/api.domain.com$fastcgi_script_name; fastcgi_param HTTPS on; include fastcgi_params; } location ~ /\.ht { deny all; } } nginx.conf user nginx; worker_processes 1; error_log /var/log/nginx/error.log; pid /var/run/nginx.pid; events { worker_connections 1024; } http { include /etc/nginx/mime.types; default_type application/octet-stream; log_format main '$remote_addr - $remote_user [$time_local] "$request" ' '$status $body_bytes_sent "$http_referer" ' '"$http_user_agent" "$http_x_forwarded_for"'; access_log /var/log/nginx/access.log main; sendfile on; #tcp_nopush on; #keepalive_timeout 0; keepalive_timeout 65; gzip on; include /etc/nginx/conf.d/*.conf; } I have a server running on port 80 that runs with no issues. As soon as I turn on this api server running on ssl, it works for about a minute and then crashes and gives a 504 Gateway Time-out. Running nginx/1.2.3

    Read the article

  • Priorize /server-status front another .htaccess rewrite rules in apache2

    - by Wiliam
    It's possible to priorize /server-status front another rewrite rules? For example the next ones don't let the /server-status works: RewriteEngine On RewriteCond %{REQUEST_FILENAME} !-s RewriteCond %{REQUEST_FILENAME} !-l RewriteCond %{REQUEST_FILENAME} !-d RewriteCond %{REQUEST_URI} !\. RewriteRule ^.*$ index.php [NC,L] I have to add this to the .htaccess to don't enter to the php project: RewriteRule ^server-status$ - [L] With mod_pagespeed of google, their statistics page works without adding them to the project .htaccess: /mod_pagespeed_statistics https://developers.google.com/speed/docs/mod_pagespeed/configuration My mod_status.conf: <IfModule mod_status.c> <Location /server-status> SetHandler server-status Order Deny,Allow Deny from all Allow from 127.0.0.1 # My allowed ips here </Location> </IfModule>

    Read the article

  • Detecting Request that uses invalid Encoding using Modsecurity

    - by Ali Ahmad
    I am trying write a virtual patch using modsecurity for my hosted web application using following rule i.e. <Location /index.php> SecDefaultAction phase:2,t:none,log,deny # Validate parameter names SecRule ARGS_NAMES "!^(articleid)$" \ "msg:'Unknown parameter: %{MATCHED_VAR_NAME}'" # Expecting articleid only once SecRule &ARGS:articleid "!@eq 1" \ "msg:'Parameter articleid seen more than once'" # Validate parameter articleid SecRule ARGS:articleid "!^[0-9]{1,10}$" \ "msg:'Invalid parameter articleid'" </Location> The problem is how can i reject requests that use invalid encoding as a global WAF configuration so that this patch cannot be circumvented.

    Read the article

  • Domains propagation issues.

    - by Andrew
    Hello to all. I got very strange issue, really weird. On weekend, May 9th I changed my server location from US to UK. Of course, everything works correctly excluding domains. There's something wrong. I got few domains on this server but I still cannot access them. When I try from the other location it works correctly. The most funny situation is that everything is working correctly from my girlfriend's work, about 500 meters from our house, but they have another ISP. It also works when I access the domains via proxy server. I checked who.is informations and everything seems to be working. On Sunday and today morning I was able to access my domains but only for a while. When I refreshed website second time I got error "Firefox was not able to connect server". Since then I'm still getting this error. Could it be my ISP fault? Regards, Andrew

    Read the article

  • Javacing code in terminal havivng a Jar in CLASSPATH

    - by Masi
    How can you javac the code in terminal by using google-collections in CLASSPATH? Example of code trying to javac in terminal (works in Eclipse) import com.google.common.collect.BiMap; import com.google.common.collect.HashBiMap; public class Locate { ... BiMap<MyFile, Integer> rankingToResult = HashBiMap.create(); ... } Javacing in terminaling src 288 % javac Locate.java Locate.java:14: package com.google.common.collect does not exist import com.google.common.collect.BiMap; ^ Locate.java:15: package com.google.common.collect does not exist import com.google.common.collect.HashBiMap; ^ Locate.java:153: cannot find symbol symbol : class BiMap location: class Locate BiMap<MyFile, Integer> rankingToResult = HashBiMap.create(); ^ Locate.java:153: cannot find symbol symbol : variable HashBiMap location: class Locate BiMap<MyFile, Integer> rankingToResult = HashBiMap.create(); ^ 4 errors My CLASSPATH src 289 % echo $CLASSPATH /u/1/bin/javaLibraries/google-collect-1.0.jar

    Read the article

  • Who moved this PC here aka Locate which port on switch a PC is plugged into

    - by ggonsalv
    The catch is you have no SNMP access, not even public. The end vision is locate a PC in building easily even if PC's are moved around. The MAC address of the PC is known and the software would run as client on each desktop, reporting back which port the PC was plugged into. Well from a programmer perspective, my network skills are not the best. Yes I could use SNMP, download the MAC port table, load it into SQL, match it to the PC name. Seems alot of work. Lets say I ping a single point from the PC. Would the echo have some thing unique for each device on the same switch? All I need to identify some thing unique for each PC plugged into each port. If the PC was moved from location A to a different location then the unique response would change.

    Read the article

  • Nginx: Rewriting entire URI to query string

    - by Doug
    Still pretty new to nginx here, trying to get a simple rewrite to work, but the server just responds '404 not found' My server block server { listen 80; listen [::]:80; server_name pics.example.com; root /home/pics; rewrite ^(.*)$ index.php?tag=$1; location / { try_files $uri $uri/ $uri.php /index.html $uri =404; #try_files $uri =404; fastcgi_split_path_info ^([a-z]+)(/.+)$; include fastcgi_params; fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name; fastcgi_pass unix:/tmp/php5-fpm.sock; fastcgi_index index.php; } location /doc/ { alias /usr/share/doc/; autoindex on; allow 127.0.0.1; deny all; } } pics.example.com/foobear should rewrite to pics.example.com/index.php?tag=foobear

    Read the article

  • Where does Windows 8 put the exe of the default browser for Modern UI?

    - by avirk
    I was trying to some hack with Win-8 and I found something which is really gonna out my mind. When I set the default browser to the IE then its icon become Modern UI and I can't see the option Open file location at the bottom when I select it by right click. But if I don't set it default then it became a desktop version icon and show up the option when I select it. Same is for the Google Chrome when I checked it. IE icon when it is not set to default, I can see the option open file location. Google Chrome icon when it is set to default. Google Chrome of desktop version when it is set to default. So my question is where does Modern UI keep exe of the default browser? And why the default browser has Modern UI icon and non-default browser has desktop version icon.

    Read the article

  • Setting up nginx as proxy to apache; All good, but nginx doesn't serve media

    - by becomingGuru
    I have set it up such that nginx proxies request and sends django requests to apache and serves media itself. Following documents my setup: Nginx Configuration: /etc/nginx/nginx.conf user www-data; worker_processes 1; error_log /var/log/nginx/error.log; pid /var/run/nginx.pid; events { worker_connections 1024; } http { include /etc/nginx/mime.types; default_type application/octet-stream; access_log /var/log/nginx/access.log; sendfile on; tcp_nopush on; #keepalive_timeout 0; keepalive_timeout 65; tcp_nodelay on; gzip on; include /etc/nginx/sites-enabled/*; } ===== ngnix proxy /etc/nginx/proxy.conf ============ proxy_redirect off; proxy_set_header Host $host; proxy_set_header X-Real-IP $remote_addr; proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; client_max_body_size 10m; client_body_buffer_size 128k; proxy_connect_timeout 90; proxy_send_timeout 90; proxy_read_timeout 90; proxy_buffer_size 4k; proxy_buffers 4 32k; proxy_busy_buffers_size 64k; proxy_temp_file_write_size 64k; =========== Nginx server file: /etc/nginx/sites-enabled/some-name.txt ========== server { listen 208.109.252.110:80; server_name netconf; autoindex on; access_log /home/site/server_logs/nginx_access.log; error_log /home/site/server_logs/nginx_error.log; location / { proxy_pass http://127.0.0.1:80/; include /etc/nginx/proxy.conf; } location /site_media/ { root /home/site/folder/static; } } ========== Nginx very well proxies the request and passes to apache, the required requests, but doesn't serve the media. In the last server file, location site_media is not served, at all. :( Everything seems perfect to me. What is wrong? Thanks in advance.

    Read the article

  • trying to install lync 2010 and experiecing an error with central management store

    - by Itai Ganot
    I'm trying to install Lync 2010 and i'm getting stuck in the stage where i have to install or point to the local configuration store. I've tried finding it in the domain and without luck, any recommendations? PS C:\Users\Administrator.ASUTA> Get-csconfigurationstorelocation WARNING: No Configuration Store location has been set. PS C:\Users\Administrator.ASUTA> Get-CsComputer "$env:computername.$env:userdnsd omain" Get-CsComputer : Cannot find location of Central Management Store in Active Dir ectory. At line:1 char:15 + Get-CsComputer <<<< "$env:computername.$env:userdnsdomain" + CategoryInfo : ResourceUnavailable: (:) [Get-CsComputer], Manag ementStoreNotFoundException + FullyQualifiedErrorId : ManagementStoreNotFound,Microsoft.Rtc.Management .Xds.GetComputerCmdlet PS C:\Users\Administrator.ASUTA>

    Read the article

  • Second DocumentRoot for certain URLS

    - by scrr
    Hello, I have the following setup in my apache-config: <VirtualHost 1.2.3.4:80> ServerName example.com:80 ServerAlias www.example.com DocumentRoot /var/www/page <Location "/blog"> DocumentRoot /var/www/blog </Location> RailsBaseURI / RailsEnv development </VirtualHost> However, Apache tells me, I am not allowed to have a second DocumentRoot. How can I make "www.example.com/blog" point to "/var/www/blog"? I'm sure this is basic, but I just can't find the proper documentation online.

    Read the article

  • CentOS 6.5 SVN https - Unknown DAV provider: svn

    - by Programster
    I am trying to setup a CentOS 6.5 64bit server with SVN over HTTPS. Unfortunately after configuring the /etc/httpd/conf.d/subversion.conf file as follows (changed paths): <Location /repos> DAV svn SVNParentPath /path/to/svn/repos # Limit write permission to list of valid users <LimitExcept GET PROPFIND OPTIONS REPORT> # Require SSL connection for password protection SSLRequireSSL AuthType Basic AuthName "Authorization Realm" AuthUserFile /path/to/passwdfile Require valid-user </LimitExcept> </Location> I get the following error message when restarting http: Starting httpd: Syntax error on line 3 of /etc/httpd/conf.d/subversion.conf: Unknown DAV provider: svn I have triple checked that I have the mod_dav_svn package already installed: Package mod_dav_svn-1.6.11-10.el6_5.x86_64 already installed and latest version Is my config wrong or are there other packages I need to set up?

    Read the article

  • Unable to find valid certification path to requested target while CAS authentication

    - by Dmitriy Sukharev
    I'm trying to configure CAS authentication. It requires both CAS and client application to use HTTPS protocol. Unfortunately we should use self-signed certificate (with CN that doesn't have anything in common with our server). Also the server is behind firewall and we have only two ports (ssh and https) visible. As far as there're several application that should be visible externally, we use Apache for ajp reverse proxying requests to these applications. Secure connections are managed by Apache, and all Tomcat are not configured to work with SSL. But I obtained exception while authentication, therefore desided to set keystore in CATALINA_OPTS: export CATALINA_OPTS="-Djavax.net.ssl.keyStore=/path/to/tomcat/ssl/cert.pfx -Djavax.net.ssl.keyStoreType=PKCS12 -Djavax.net.ssl.keyStorePassword=password -Djavax.net.ssl.keyAlias=alias -Djavax.net.debug=ssl" cert.pfx was obtained from certificate and key that are used by Apache HTTP Server: $ openssl pkcs12 -export -out /path/to/tomcat/ssl/cert.pfx -inkey /path/to/apache2/ssl/server-key.pem -in /path/to/apache2/ssl/server-cert.pem When I try to authenticate a user I obtain the following exception: Caused by: sun.security.provider.certpath.SunCertPathBuilderException: unable to find valid certification path to requested target at sun.security.provider.certpath.SunCertPathBuilder.engineBuild(SunCertPathBuilder.java:174) ~[na:1.6.0_32] at java.security.cert.CertPathBuilder.build(CertPathBuilder.java:238) ~[na:1.6.0_32] at sun.security.validator.PKIXValidator.doBuild(PKIXValidator.java:318) ~[na:1.6.0_32] Meanwhile I can see in catalina.out that Tomcat see certificate in cert.pfx and it's the same as the one that is used while authentication: 09:11:38.886 [http-bio-8080-exec-2] DEBUG o.j.c.c.v.Cas20ProxyTicketValidator - Constructing validation url: https://external-ip/cas/proxyValidate?pgtUrl=https%3A%2F%2Fexternal-ip%2Fclient%2Fj_spring_cas_security_proxyreceptor&ticket=ST-17-PN26WtdsZqNmpUBS59RC-cas&service=https%3A%2F%2Fexternal-ip%2Fclient%2Fj_spring_cas_security_check 09:11:38.886 [http-bio-8080-exec-2] DEBUG o.j.c.c.v.Cas20ProxyTicketValidator - Retrieving response from server. keyStore is : /path/to/tomcat/ssl/cert.pfx keyStore type is : PKCS12 keyStore provider is : init keystore init keymanager of type SunX509 *** found key for : 1 chain [0] = [ [ Version: V1 Subject: CN=wrong.domain.name, O=Our organization, L=Location, ST=State, C=Country Signature Algorithm: SHA1withRSA, OID = 1.2.840.113549.1.1.5 Key: Sun RSA public key, 1024 bits modulus: 13??a lot of digits here??19 public exponent: ????7 Validity: [From: Tue Apr 24 16:32:18 CEST 2012, To: Wed Apr 24 16:32:18 CEST 2013] Issuer: CN=wrong.domain.name, O=Our organization, L=Location, ST=State, C=Country SerialNumber: [ d??????? ????????] ] Algorithm: [SHA1withRSA] Signature: 0000: 65 Signature is here 0070: 96 . ] *** trustStore is: /jdk-home-folder/jre/lib/security/cacerts Here is a lot of trusted CAs. Here is nothing related to our certicate or our (not trusted) CA. ... 09:11:39.731 [http-bio-8080-exec-4] DEBUG o.j.c.c.v.Cas20ProxyTicketValidator - Retrieving response from server. Allow unsafe renegotiation: false Allow legacy hello messages: true Is initial handshake: true Is secure renegotiation: false %% No cached client session *** ClientHello, TLSv1 RandomCookie: GMT: 1347433643 bytes = { 63, 239, 180, 32, 103, 140, 83, 7, 109, 149, 177, 80, 223, 79, 243, 244, 60, 191, 124, 139, 108, 5, 122, 238, 146, 1, 54, 218 } Session ID: {} Cipher Suites: [SSL_RSA_WITH_RC4_128_MD5, SSL_RSA_WITH_RC4_128_SHA, TLS_RSA_WITH_AES_128_CBC_SHA, TLS_DHE_RSA_WITH_AES_128_CBC_SHA, TLS_DHE_DSS_WITH_AES_128_CBC_SHA, SSL_RSA_WITH_3DES_EDE_CBC_SHA, SSL_DHE_RSA_WITH_3DES_EDE_CBC_SHA, SSL_DHE_DSS_WITH_3DES_EDE_CBC_SHA, SSL_RSA_WITH_DES_CBC_SHA, SSL_DHE_RSA_WITH_DES_CBC_SHA, SSL_DHE_DSS_WITH_DES_CBC_SHA, SSL_RSA_EXPORT_WITH_RC4_40_MD5, SSL_RSA_EXPORT_WITH_DES40_CBC_SHA, SSL_DHE_RSA_EXPORT_WITH_DES40_CBC_SHA, SSL_DHE_DSS_EXPORT_WITH_DES40_CBC_SHA, TLS_EMPTY_RENEGOTIATION_INFO_SCSV] Compression Methods: { 0 } *** http-bio-8080-exec-4, WRITE: TLSv1 Handshake, length = 75 http-bio-8080-exec-4, WRITE: SSLv2 client hello message, length = 101 http-bio-8080-exec-4, READ: TLSv1 Handshake, length = 81 *** ServerHello, TLSv1 RandomCookie: GMT: 1347433643 bytes = { 145, 237, 232, 63, 240, 104, 234, 201, 148, 235, 12, 222, 60, 75, 174, 0, 103, 38, 196, 181, 27, 226, 243, 61, 34, 7, 107, 72 } Session ID: {79, 202, 117, 79, 130, 216, 168, 38, 68, 29, 182, 82, 16, 25, 251, 66, 93, 108, 49, 133, 92, 108, 198, 23, 120, 120, 135, 151, 15, 13, 199, 87} Cipher Suite: SSL_RSA_WITH_RC4_128_SHA Compression Method: 0 Extension renegotiation_info, renegotiated_connection: <empty> *** %% Created: [Session-2, SSL_RSA_WITH_RC4_128_SHA] ** SSL_RSA_WITH_RC4_128_SHA http-bio-8080-exec-4, READ: TLSv1 Handshake, length = 609 *** Certificate chain chain [0] = [ [ Version: V1 Subject: CN=wrong.domain.name, O=Our organization, L=Location, ST=State, C=Country Signature Algorithm: SHA1withRSA, OID = 1.2.840.113549.1.1.5 Key: Sun RSA public key, 1024 bits modulus: 13??a lot of digits here??19 public exponent: ????7 Validity: [From: Tue Apr 24 16:32:18 CEST 2012, To: Wed Apr 24 16:32:18 CEST 2013] Issuer: CN=wrong.domain.name, O=Our organization, L=Location, ST=State, C=Country SerialNumber: [ d??????? ????????] ] Algorithm: [SHA1withRSA] Signature: 0000: 65 Signature is here 0070: 96 . ] *** http-bio-8080-exec-4, SEND TLSv1 ALERT: fatal, description = certificate_unknown http-bio-8080-exec-4, WRITE: TLSv1 Alert, length = 2 http-bio-8080-exec-4, called closeSocket() http-bio-8080-exec-4, handling exception: javax.net.ssl.SSLHandshakeException: sun.security.validator.ValidatorException: PKIX path building failed: sun.security.provider.certpath.SunCertPathBuilderException: unable to find valid certification path to requested target I tried to convert our pem certificate to der format and imported it to trustedKeyStore (cacerts) (without private key), but it didn't change anything. But I'm not confident that I did it rigth. Also I must inform you that I don't know passphrase for our servier-key.pem file, and probably it differs from password for keystore created by me. OS: CentOS 6.2 Architecture: x64 Tomcat version: 7 Apache HTTP Server version: 2.4 Is there any way to make Tomcat accepts our certificate?

    Read the article

  • Serving static files fails - nginx

    - by Sergei
    Hi, I've been looking and trying around all night, but without success. I configured nginx to serve my static files and proxy all the other traffic: server { listen 80; server_name mydomain.com; access_log /home/boudewijn/www/bbt/brouwers/logs/access.log; error_log /home/boudewijn/www/bbt/brouwers/logs/error.log; location / { proxy_pass http://127.0.0.1:8080; include /etc/nginx/proxy.conf; } location /media/ { root /home/boudewijn/www/bbt/brouwers/; } } The proxy passing is no problem, but when I go to mydomain.com/media/ or try to access any testfile over there, it's without success. I paid attention to the difference between root and alias, my media folder exists, I paid attention to the trailing slashes, but still I get a 404 when trying to access my static media files. Any help?

    Read the article

  • Disable public Tomcat6 stack trace

    - by The NinjaSysadmin
    Can anyone advise me how I can go about disabling Tomcat6 from displaying stacktrace output to the browser? Tomcat: 6.0.29 I have made the following changes to /opt/apache-tomcat-6.0.29/conf/web.xml <error-page> <exception-type>java.lang.Throwable</exception-type> <location>/error.jsp</location> </error-page> I'm told putting this in place will give a white screen if the file doesn't exist, however I'm getting stack traces to the screen.

    Read the article

  • Enable Server Status using Plesk 11

    - by Lars Ebert
    I am trying to get apaches server status to work with Plesk 11. But running sudo /usr/sbin/apache2ctl fullstatus results in: Forbidden You don't have permission to access /server-status on this server. __________________________________________________________________ Apache Server at localhost Port 80 'www-browser -dump http://localhost:80/server-status' failed. Maybe you need to install a package providing www-browser or you need to adjust the APACHE_LYNX variable in /etc/apache2/envvars How can I enable server status? So far I have tried to insert <Location /server-status> SetHandler server-status Order Deny,Allow Deny from all Allow from localhost </Location> into the httpd.conf, but I am not sure if it is active. I also tried adding it to /var/www/vhosts/somedomain/conf/vhost.conf but I do not know which domain I have to add this to, as fullstatus seems to query localhost directly. I guess I am a little confused by the use of vhost configuration in Plesk.

    Read the article

< Previous Page | 99 100 101 102 103 104 105 106 107 108 109 110  | Next Page >