The following is a list of how I solved a large variety of issues encountered in the last 5+ years as a web developer. In many cases, I shorten down the problem and solution into 2 lines, even though the solution may have taken many, many hours to uncover.
Keep in mind the problems and solutions may not match your situation exactly. Use at your own risk.
Hopefully this thread can save a few folks hours of their lives, thus paying it forward for all the Google search help over the years.
Problem: ‘Missing Dependency: perl-Git’ error when trying to install git on a cPanel server
Solution: yum --disableexcludes=main install --enablerepo=webtatic git , where webtatic is an already installed repository
Problem: git submodule update --init fails with No submodule mapping found in .gitmodules for path 'somedirectory'
Solution: The directory ‘somedirectory’ already exists. The git submodule is trying to initialize the directory, but can’t because it already exists. Do an rmdir somedir and git rm somedir and commit, push, and then re-run git submodule update --init
Problem: github.com[0: 207.97.227.239]: errno=Connection timed out fatal: unable to connect a socket (Connection timed out)
Solution: Open port 9418 outbound in any firewalls.
Problem: mongod dead but subsys locked
Solution: rm -f /var/lock/subsys/mongod && rm -f /var/lib/mongodb/mongod.lock
Problem: protocol version mismatch - is your shell clean?" when trying to do an rsync over SSH using authorized_keys
Solution: Change the shell from /sbin/nologin to a shell that allows logging in via SSH
Problem: When I try to ssh <hostname> I get an error " name or service not known"
Solution: `pico /etc/resolv.conf` and check for any bad DNS servers. For instance, this issue occurred because an old local IP was in the file after the server was moved to a new IP
Problem: Running ssh -v shows the error message 'Unspecified GSS failure'
Solution: Disable GSSAPI Authentication by setting "GSSAPIAuthentication no" in /etc/ssh/sshd_config
Problem: Public Authentication doesn't work and /var/log/auth or /var/log/secure show "Authentication refused: bad ownership or modes for directory /home/USERNAME"
Solution: Check that the permissions on the /home/USERNAME folder don't allow for write access to group, so something like 650
Problem: Public Authentication doesn't work after manually creating a .ssh directory and .ssh/authorized_keys and adding a key
Solution: Check that permissions on the .ssh dir is 700 and the .ssh/authorized_keys is 600
Problem: SSH works, but SFTP connection does not. /var/log/secure shows "subsystem request for sftp"
Solution: In /etc/ssh/sshd_config, set Subsystem sftp internal-sftp
Per https://forums.gentoo.org/viewtopic-t-733616.html, In a chrooted environment, /usr/ is unlikely to exist. That's alright though because you don't actually need to execute /usr/lib/misc/sftp-server to handle an SFTP connection. The SSH daemon has this functionality built-in.
Problem: 'bad interpreter permission denied' AND 'cannot run C compiled programs' while trying to ./configure subversion
Solution: Move the build source out of the /tmp/ directory and into /home/username
Problem: While trying to compile and make the PHP PECL Extension for svn an error while running `make` was "/usr/bin/ld: cannot find -lsvn_client-1 collect2: ld returned 1 exit status make: *** [svn.la] Error 1"
Solution: Opened the Makefile and changed it to point at the folder where the libsvn_client libraries, and other SVN libraries, were located. SVN_SHARED_LIBADD = -Wl,-rpath,/usr/local/subversion/lib -L/usr/local/subversion/lib -lsvn_client-1 -lsvn_fs-1 -lsvn_repos-1 -lsvn_subr-1 -lapr-1 it was point to /usr/local/apache/lib
Problem: Error "Disk quota exceeded", but the disk space was at 44% and no quotas were set. Attempting to add one more file caused the error. Deleting any file, even a larger file, would allow for one file to replace it.
Solution: Hosting company had to increase the inodes limit for the VPS. This limits the number of files on a filesystem.
Problem: /usr/sbin/mingetty is eating up CPU
Solutions: As root, mv /var/run/utmp /var/run/utmp.BAK && touch /var/run/utmp
Problem: When running yum updates “TypeProblem: unsubscriptable object”
Solution: yum clean metadata
Problem: Can't connect to MySQL server on 'mysql.example.com' (13).
Solution: SELinux blocking, run `setsebool -P httpd_can_network_connect=1`
Problem: #1037 - Out of memory; restart server and try again -- generated while running a query with ORDER BY. It became an issue after Furman and Britt Peters' servers were migrated to new machines.
Solution: It seems this issue can be fixed by increasing the "sort_buffer_size" memory value in MySQL's /etc/my.cnf configuration file. Increase to 80M for Furman before it worked. This is likely too high, but it worked.
Problem: #1114 - The table 'COLUMNS' is full OR The table '#sql-1cb8_39a1' is full query (while doing an ALTER ADD INDEX
Solution 1: Check the /etc/my.cnf and test increasing or decreasin limits on certain variables. This fixed the issue when an old my.cnf with high memory values was showing "table full"
Solution 2: Most likely there are odd character set characters in the data. For instance, if the database is old and was created on a Windows computer and moved to Linux. Try identifying odd dashes, hypens, ascent marks, apostrophes and double_quotes and replace them like UPDATE node_revisions SET body = REPLACE(body, "’", "'")
Solution 3: Bug in version 5.0.45, upgrade MySQL. http://bugs.mysql.com/bug.php?id=31379
Problem: MySQL stops working. "ERROR! MySQL manager or server PID file could not be found! Starting MySQL. ERROR! Manager of pid-file quit without updating file."
Reason: This occurred when the hostname of the server was changed. A process ID file using the hostname must get changed in the process and causes theProblem.
Solution: First, make sure your running the command as the root user or via sudo. Use `top` to find the PID of mysqld_safe and/or mysqld Then use `pkill -P {pid}` to stop the existing proceess(es). Then do a `sudo /etc/init.d/mysql start`
This may also occur if the diskinodes getting full, at which point the host would need to double the existing limit on the server
Problem: MySQL won't start. "ERROR! Manager of pid-file quit without updating file."
Solution: First, make sure your running the command as the root user or via sudo. Check for errors in the configuration file /etc/my.cnf
Problem: Mysql doesn't start on a server reboot.
Solution: Enable mysql or mysql in run-levels 3,4,5 sudo /sbin/chkconfig --level 345 mysqld on
To check existing run-levels to sudo /sbin/chkconfig --list if mysql or mysqld doesn't exist then do making sure there exists a /etc/init.d/mysql (it must be executable – chmod +x), doing a chkconfig --add mysql and making sure that it starts up at the correct runlevels (typically 3, 4 and 5). A simple chkconfig --level 345 mysql on and you’re ready to rock.
Problem: Wrong bytesec when repairing a table (note: this was reached when reached during a test of restoring a database by doing a schema dump, moving .MYD files, and repairing tables/indexes)
Solution: In this case the wrong .MDY file was copied into the restore. (Note that this method of restoring is not 100% trusted, it was just a test)
Problem: When trying to import a database through PHPMyAdmin, “1153 -- 'max_allowed_packet' exceeded”
Solution: Under the [mysqld] section of /etc/my.cnf add a value like max_allowed_packet = 50M
Problem: When saving the CSF firewall configuration through WHM --- Problem: The VPS iptables rule limit (numiptent) is too low (119/128) - stopping firewall to prevent iptables blocking all connections, at line 1006"
Solution: Have the hosting company set the numiptent value has been set to 256 on your server and are able to restart the firewall The suggestedSolution on many message boards it that the VPS hosting company is able to increase the numiptent value.
Problem: Unrecognized command line option
Solution: http://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg75066.html
Problem: Brute force error message.
Solution: Login via SSH. run mysql -u root -p and login to MySQL. SELECT cphulkd; delete from brutes; delete from logins;
Problem: "Invalid License File" message upon logging into WHM
Solution: Connect via SSH and issue the command `service cpanel restart` alternatively `/usr/local/cpanel/cpkeyclt` might also be necessary
Problem: Free IP addresses not showing under “Change Site IP Address” interface
Solution: Run ‘Rebuild the IP address Pool’ from the sidebar
Problem: In cPanel Cron Jobs section "/usr/bin/crontab permissions are wrong"
Solution: `chmod 4755 /usr/bin/crontab`
Problem: Error "Error fetching zone data for cpanel" when trying to edit the MX records.
Solution: This was a result of deleting the zone for the main domain. Went to "Add a DSN Zone" and then add a new zone. Then return to cPanel and edit the MX records as usual under Mail >> MX Entry
Problem: Error message sent to email "syslogd failed @ Fri Feb 19 16:01:01 2010. A restart was attempted automagically."
Solution: Restart the log daemon /etc/init.d/rsyslog restart
Problem: In WHM "Your license could not be activated because: Unactivated License File"
Solution: Contact the hosting company if the cPanel site also says the IP address isn't registered. If it says the IP is registered then run
Problem: When running /scripts/upcp “Undefined subroutine &Cpanel::Update::automatic_updates_enabled called at /scripts/upcp line 273.”
Solution: http://forums.cpanel.net/f5/scripts-upcp-force-produces-error-173992.html
Problem: Log \#015: invalid numeric value error
Solution: Don't create initial scripts in Windows because this adds hidden carriage returns (i.e. \r or character #015)
Problem: While executing an sh script the error "command not foundpts" and "not a valid identifier" along with strange behavior
Solution: Character set of the file is screwed up. Delete the file and create a new copy from scratch. The text editor used to create the original may be inserting breaklines or other non standard characters.
Named/bind
Problem: Error starting bind/named - "Starting named: /usr/sbin/named: symbol lookup Problem: /usr/sbin/named: undefined symbol: dns_dispatch_hash"
Solution: Reinstall bind, using a "cPanel >> Upgrade to Latest Version" and the "Service Configuration >> Nameserver Setup" in WHM/cPanel
Problem: While installing SSL Cert - in Firefox got the message "SSL received a record that exceeded the maximum permissible length. (Error code: ssl_error_rx_record_too_long)"
Solution: Apache configuration was missing SSLEngine On
Problem: In Apache error_log - SoftException in Application.cpp:376: Mismatch between target GID (32013) and GID (99)
Solution: Use chmod to set the correct permissions on the directory, like chmod 32011:32013 * and chmod 32011:32013 .*
Problem: In Apache error_log - pcfg_openfile: unable to check htaccess file, ensure it is readable
Solution: chmod 755 on public_html
Problem: 404 error using mod_userdir /~username and Drupal
Solution: Set the RewriteBase directive in .htaccess to /~username
Problem: On apache2 restart "Syntax error on line 300 of /etc/apache2/apache2.conf: Cannot load /usr/local/apache2/modules/mod_dav_svn.so into server: /usr/local/subversion/lib/libsvn_subr-1.so.0: undefined symbol: apr_hash_clear"
Solution: Recompile SVN for Apache Server
Problem: On apache2 restart "/usr/local/apache2/conf/httpd.conf: Cannot load /usr/local/apache2/modules/mod_dav_svn.so into server: /usr/lib/libsvn_fs_base-1.so.0: undefined symbol: db_create "
Solution: Install subversion 1.5.5 from source, link to the .so it creates and restart began working. Possibly an issue in the yum build not lining to BerkleyDB and/or BDB not being installed
Problem: After using EasyApache to rebuild Apache and PHP the error "/usr/local/modules/mod_dav_svn.so "Cannot open shared object file: No such file or directory"
Solution: SVN must be built against the new installation. Temporarily commented out the LoadModule modules/mod_dav_svn.so and LoadModule module/mod_authz_svn.so and then rebuilt SVN against Apache from source
Problem: 403 Forbidden
Solution: Ensure the linux account under which Apache is running (use `top` to find out) has access to the /home/account/public_html and also the files within it.
Problem: After installing SSL certificate, a restart shows no errors but still no connection
Solution: Verify port 443 is open. Open on any software (Lokkit/iptables) and hardware firewalls
Problem: 404 due to Rewrite rules and entire .htaccess file being ignored.
Solution: Enable AllOverride All via the appropriate <Directory ....> setting in httpd.conf
Problem: On Apache restart, “DocumentRoot /home/someuser/public_html does not exist.
Solution: SELinux is blocking. Use `setsebool -P httpd_enable_homedirs true` and then change the flags on the home directory via `chcon -R -t httpd_sys_content_t /home/someusername/public_html` then restart Apache again.
Problem: On restart "mixing * ports and non-* ports with a NameVirtualHost address is not supported"
Solution: If using cPanel/WHM then try running /scripts/rebuildhttpdconf
Problem: Recompiling with suPHP errors in Apache error log "SoftException in Application.cpp:592: Directory / is not owned by alliance" and "Premature end of script headers: index.php"
Solution: suPHP requires the / directory be owned by root. It was owned by a random userid. chown root:root / fixed this issue
Problem: After recompiling with suPHP errors in Apache error log “SoftException in Application.cpp:256: File "/home/example/public_html/index.php" is writeable by group”
Solution: Remove group write permission chmod -Rf g-w *
Problem: Configuring SSL Certificate in Apache .conf files with a .crt and .pem says “does not exist or is empty”
Solution: SELinux can’t read the files. Update the context to be readable by httpd like sudo chcon -R -t httpd_user_content_t somfilename.crt
Problem: On Apache restart Starting httpd: (98)Address already in use: make_sock: could not bind to address 0.0.0.0:80 no listening sockets available, shutting down Unable to open logs
Solution: fuser -k -n tcp 80 (which kills any processes running on port 80
Problem: In Apache error logs, the following appears “symbolic link not allowed or link target not accessible” and files that are located in a symlink directory are not accessible.
Solution: In the /etc/http/conf/httpd.conf the main Options line probably includes +SymLinksIfOwnerMatch, it will need to be -SymLinksIfOwnerMatch in order for a user to follow a symlink outside of their home directory. Or, if using cPanel WHM go to Main >> Service Configuration >> Apache Configuration >> Global Configuration Main >> Service Configuration >> Apache Configuration >> Global Configuration and uncheck the SymLinksIfOwnerMatch checkbox and save.
Drush
Problem: When trying to do a drush update -l example.com Warning: Cannot use a scalar value as an array in ~/public_html//sites/all/modules/domain/domain.bootstrap.inc on line 58 Warning: session_start(): Cannot send session cookie - headers already sent by (output started at ~/public_html/sites/all/modules/domain/domain.bootstrap.inc:58) in ~/public_html//includes/bootstrap.inc on line 1031
Solution: If the domains module is install the include statement at the bottom of the settings.php must be temporarily commented out.
Problem: HTTP request staus fails, or Cron run failed, or Unable to fetch any information about available new releases and updates.
Solution: Correct Nameservers so they resolve the external IP addresses these scripts are calling. More details at http://drupal.org/node/222454#comment-1136650
Problem: httpd: bad user name key when trying to (re)start
Solution: Check configuration files for old VirtualHosts which declare a User or Group parameter. For instance, if an Apache conf file had a rule for a domain which is no longer hosted on the server and the account and directory have since been deleted.
Problem: “Trying to clone an uncloneable object of class mysqli” when trying to use a mysqi connection.
Solution: Adding 'php_flag zend.ze1_compatibility_mode off' to the .htaccess fixed this, plus an issue with a certain page rendering a blank page with no error messages in PHP or Apache and a HTTP/0.9 200 OK response.
Problem: Unable to login, no database access
Solution: Edit the /modules/user/user.module and bypass the validation function, like
function user_login_final_validate($form, &$form_state) {
global $user;
$user->uid = 1; // adding this line fakes out the login
Problem: On the /admin/reports/status the error "HTTP request status fails" appears
Solution: One possibleSolution is to change run the query UPDATE variable SET value = 'b:0' WHERE name = 'drupal_http_request_fails';
Problem: While upgrading from Drupal 5 to 6, a copy of the live Drupal database was made and setup as a development. When trying to login to Drupal (/user) with the development db Drupal would open a session but show "not authorized to access this page" on any secured page.
Solution: Restart MySQL
Problem: "Attempting to re-run cron while it is already running."
Solution: Clear the cache and remove cron_semaphore and cron_last from the variables table. If the cron job gets "stuck" and you're unable to access any admin/* page then just wait a while. It seems the php job may need to timeout.
Problem: After migrating a D6 sandbox to a new server the /admin/* pages, but not regular pages showed "maximum execution time of 30 seconds exceeded" in various /include/* files.
Solution: Set php_value max_execution_time 90 in the .htaccess file
Problem: The Comment wasn't showing a preview when the "Preview Comment" button was pressed
Solution: The button value was changed, it must be labedl "Preview", instead of something custom like "Preview Comment" because the comment_form function checks for ($op == t('Preview')
Problem: "directory does not exist" or "directory is not writable", despite 777 and appropriate permissions
Solution: On Fedora and other distros the SELinux can be running. Set each directory to make writeable by httpd, like this example
sudo /usr/sbin/semanage fcontext -a -t httpd_sys_content_t "/home/user/public_html(/.*)?"
sudo /sbin/restorecon -R ~/public_html
To temporarily disable run setenforce 0 See http://drupal.org/node/50280 for more info
To set SELinux to only log errors edit /etc/selinux/config and set SELINUX=permissive (as described in man selinux)
Problem: In Drupal, errors "Error sending e-mail". In Sendmail log "NOQUEUE: SYSERR /etc/mail/sendmail.cf Permission denied". PHP's mail function works from the command line, but not via the webserver.
Solution: SELinx is blocking httpd from sending mail. To fix, run Then run sudo /usr/sbin/semanage boolean -l|grep sendmail and verify that the output of the command shows httpd_can_sendmail -> on httpd_can_sendmail
Note: Without the -P option, only the current boolean value is affected; the boot-time default settings are not changed. If the -P option is given, all pending values are written to the policy file on disk. So they will be persistent across reboots.
Problem: Ubercart's Payment methods "pane" doesn't render
Solution: Ajax was trying to call back to pages like, http://example.org/?q=cart/checkout/payment_details/credit, but these pages needed to use the https version. Adding /cart/checkout/* to the Securepages settings /admin/settings/securepages allowed the https versions of these ajax pages to load.
Problem: Unable to login because visiting /user redirects to https:// and no SSL certificate is present (usually when moving a copy to a development server or a dev domain name)
Solution: Temporarily move the /sites/all/modules/securepages directory to another name like /sites/all/modules/temp-securepages
Problem: Fatal Problem: Unsupported operand types in ~/public_html/flex-tek/modules/block/block.admin.inc on line 39 when accessing /admin/build/block
Solution: Occurred when a theme was renamed. The old theme name was still being set as the default in a settings.php. Removing it ultimately fixed the issue. However, as best practice, before moving or deleting a theme you should really disable any blocks associated to it, and uncheck any options associated with it on /admin/build/theme
Problem: Went trying to use Drush just after it was installed. includes/table.inc: No such file or directory Drush needs a copy of the PEAR Console_Table library saved as Drush [error] includes/table.inc. Drush attempted to download this automatically, but failed. To continue you will need to download the 1.1.3 package from http://pear.php.net/package/Console_Table, extract, and move the file Table.php to includes/table.inc.
Solution: Log out and log back in. This happened when I installed drush as root and tried to immediately run it
Problem "Cannot use a scalar value as an array" when trying to use Drush
Solution: This is due to the "Domains" module requiring "include './sites/all/modules/domain/settings.inc';" in the bottom of settings.php files. Temporarily comment out the "include './sites/all/modules/domain/settings.inc';", run Drush, then re-enable it.
Problem: Block(s) created in a hook_block won't appear.
Solution: Check the blocks table and see if the status flag is set to 1 without a region being set. If so, remove the record and make sure the status value is not set in the hook_block unless a region is also specified.
Problem: "MySQL server has gone away query:" message, and a lot of them, when trying to access the /admin/build/securepages page
Solution: Occurs when the 'admin/build/securepages' page is visited via http://, not https:. Use PHPMyAdmin to temporarily disable the auto switching feature by executing update variable set value = 'b:0' WHERE name Like 'securepages_switch'; Then, flush the Cache tables. Then visit https://www.example.com/admin/build/securepages (be sure to use https). Then, Add admin/build/securepages to the list of "Make secure only the listed pages." Re-enable "Switch back to http pages when there are no matches" and Save
Problem: Upon adding a new domain, content is not visible to Anonymous users, but is visible to the admin, when Domain Access module is installed.
Solution: When the Domain Strict module is enabled the roles must be explicitly granted access to use new domain. This is done under the “User Defaults” tab /admin/build/domain/roles
Problem: PHP Fatal Problem: Call to undefined function preg_last_error() in ~/public_html/sites/all/modules/invisimail/invisimail.module on line 157
Solution: Invismail uses a function that isn’t included in older version of PHP. Redhat PHP 5.1.6 doesn’t port future updates. Disabling Invismail fixes this issue.
Problem: Class 'XMLWriter' not found when trying to use xmlsitemap 6.x.2-beta2
Solution: yum install php-xml
Problem: Imagecache images won’t generate on PHP 5.1.6 even though the same code works on a different server.
Solutions: Permissions did not seem to be the issue. Going to Drupal’s Configuration -> File System and setting the download type to “Private”
was theSolution.
Problem: Error logs show “The file permissions could not be set on public://filenamehere”. and Warning: filesize() [/admin/reports/event/function.filesize">function.filesize]: stat failed for public://filenamehere in file_save() (line 573 of /home/example/public_html/includes/file.inc).
Solution: Drupal created a directory, literally named public: , in the root Drupal directory. Deleting this directory resolved the issue. This may have been caused by changing the location of the public file system path value via /admin/config/media/file-system after having used /sites/default/files Kudos for the Solution go to this thread
Problem: When running `svn --version` SVN was compiles without, and doesn’t show, handles 'https' scheme
Solution: Install devel packages `yum install expat-devel openssl-devel`
Problem: 'somedirectory' is not a working copy directory
Solution: Move that directory out of the SVN path. Run 'svn cleanup'. Move the directory back to the original location. Do an 'svn revert somedirectory'
Problem: Using Subverions + Apache2 dav_svn - error "commit failed path not found /svn/webby/!svn/9ac....3829"
Solution: The .htaccess file in the parent folder of the virtual host root was causingProblems. Probably rewrites or messing with the AuthUserFile authentication. Renamed .htaccess and it works. It's probably best not to mix an SVN repo virtualhost and a regular website with it's own .htaccess in the same folder for various security reasons.
Problem: Server sent unexpected return value (403 Forbidden) in response to PUT request for '/svn/biemedia/!svn
Solutin: Add the following within the <Location> directive
Order allow,deny
Allow from all
Problem: When using svn propedit svn:ignore . to ignore certain files, the use of svn add --force . doesn't pay attention to the ignore rules as it should according to the SVN book.
Solution: Used svn --version and found the active version was was old 1.1.4-2.ent Updated subversion to 1.5.5
Problem: When trying to use the http or https protocols the error "Unrecognized URL scheme for https"
Solution: Running svn --version shows there were no linked client libraries to handle http or https. Rebuilding the clients into the subversion and linking in neon and/or serf will provide for use of http and https
Problem: "svn: None of the environment variables SVN_EDITOR, VISUAL or EDITOR are set, and no 'editor-cmd' run-time configuration option was found" - When trying to issue a command like svn propedit svn:ignore .
Solution: Type export SVN_EDITOR=nano or for a permanent change, open ~/.bashrc and add SVN_EDITOR=nano
Problem: "error while loading shared libraries: libaprutil-1.so.0" when trying to do any svn command
Solution: Ensure the linux account is not running in a jail shell. (Go to WHM -> Manage Shell Access and change to "Enable Normal Shell"). Alternatively, there could be aProblem with the linked libraries not being read, which could be corrected by adding the .so file's path to /etc/ld.so.conf and then saving and running ldconfig
Problem: svn: Commit failed (details follow): svn: Directory '~/public_html/sites/all/modules/drush' is missing
Solution: Revert (recursively if necessary -R) the file or directory, like svn -R revert sites/all/modules/drush
Problem: Server sent unexpected return value (405 Method Not Allowed)
Solution: The file or directory already exists in the repo. Deleting the local copy and doing an SVN up may fix the issue. Then move the changes back into the working copy. If this doesn't work it may be best consider the working copy got corrupt. Make a copy of it, checkout a new copy of the entire repository and merge any changes from the corrupt copy as necessary.
Problem: When ./configure Neon “could not find library containing RSA_new”
Solution: Install openssl-devel `yum install openssl-devel` or `apt-get install openssl-devel`
Problem: When ./configure Neon “no XML parser was found: expat or libxml 2.x required”
Solution: Install expat-devel `yum install expat-devel`
Problem: svn: Checksum mismatch for
Solution: Corrupt svn files. Checkout a copy of theProblem directory to a temporary location and then replace the corrupt directory with the fresh checkout
Problem: Need to revert back to a previous version of a file or directory and commit those changes to head.
Solution: Requires a reverse merge, see http://www.jorgepablos.com/2010/03/revert-a-file-to-an-old-revision-in-subversion/
Problem: When checking out a large set of files, “svn: REPORT of '/svn/repo/!svn/vcc/default': Could not read response body: Secure connection truncated”
Solution: By default, SVNPathAutz is on in the Apache config. Disabling path-based authorization, like via mod_authz_svn, to speed up access, but since we are using mod_authz_svn, it is best to keep it on. However, having it 'on' can cause "Secure connection truncated" errors on large checkouts, due to memory leaks. In those cases it can be temporarily turned off to allow the checkout, and then turned back on See, http://svnbook.red-bean.com/en/1.5/svn.serverconfig.httpd.html
Problem: SSL handshake failed: SSL Problem: unknown protocol (https://example.com)
Solution: Ping the domain. Make sure it’s reachable.
Problem: The From: address shows Default Name <default@example.com>
Solution: The From address in the message's From: textfield should be of the form "Actual Person" <actualperson@example.com> Otherwise it may default to something like default@example.com
Problem: Email matching the web domain won't send from the web server, but email to other domains works fine.
Solution: Email is not leaving the web server and it delivered locally or to a queue. In Plesk, disabling Mail for the domain and pointing the MX records to the actual, remote MX server fixes. See Jim's post on 11/01/2010 here http://www.google.com/support/forum/p/Google+Apps/thread?tid=09165a87da6... for more details. It may be necessary to setup an A record within the web server’s DNS for the MX domain if that MX domain is the same as the emailing domain.
Lovingly crafted by orangecoat with some rights reserved, and a promise not to spam you.
Back to top