* Fix issue #1122 - Local user login fails with Multi/Single LDAP enabled
Code was not consistent with how it was generating the password hashes and comparing them. To fix this, added compat_password_xxx functions that check for the availability of the newer password_xxx functions and if not present revert back to the md5 version.
Changes made are:
auth_changepassword.php:
* Break out of switch when error message set
* Reduce duplication of code when checking passwords
* Use local variables rather than constantly pulling in request ones
* Use new compat_password_xxx functions
auth_login.php:
* Added some debug logging to identify login issues
* Allow local login even if using ldap
* Use new compat_password_xxx functions
* Remove password from SQL search as hash can change and we don't know what it would have been at the time it was set
user_admin.php
* Use new compat_password_xxx functions
lib\auth.php:
* Added compat_password_verify, compat_password_hash and compat_password_rehash to check for PHP 5.7+ compatible functions other use MD5
* Reduced SQL query to just required user in secpass_login_process()
* Fixed password check in secpass_check_history()
* Fixed return value check of secpass_check_pass() in secpass_login_process()
* Use new compat_password_xxx functions
* Update CHANGELOG
* Fixup for Issue #1122, corrects some minor issues
* Merge branch 'Issue-1122' of https://github.com/netniV/cacti into Issue-1122
Update log message to be warning notj ust a normal message
* Fix issue of local auth bypas
* Update CHANGELOG
Fix up CHANGELOG issues
* Update auth.php
Fix up formatting issues
The cacti_version_compare() function was not operating properly when
moving between second order point versions (eg 1.1.29 to 1.2.0). This
is resolved through a minor change to the cacti_version_compare()
function and the version_to_decimal() function.
Using cacti_log() with the optional POLLER_VERBOSITY_xxx paramater resulted in mising logs if they were set to the same level as the poller log level in settings
* Fix Issue #1133 - Incorrect variable name
Fixed the incorrect use of $totle in place of $title resulting in database insertion errors.
* Update import.php
Fix debug display to match code.
* Update CHANGELOG
This occurred due to the optimizations in pull request #1101. The
$sql_user must, in it's entirety, be enclosed in parenthesis.
Otherwise, expand host functionality breaks.
I've also added the site_id in this commit to reduce the effort in
attempting to maintain both the feature branch and the develop branch.
It's not used, but returned.
* Add CPU Total to 'SNMP - Get Processor Information'
This change adds a virtual index 'Total' to the SNMP info,
showing the average CPU utilization over all cores.
* Changing CPU Total in 'SNMP - Get Processor Information'
Making the total CPU utilization the first element in the array to avoid shifting when CPUs are added.
* Fix for issue #1111 - Issue with ldap authenication (#1118)
* Update change log for pull request #1111
* Fixes for issues #1084, #1106 and #1116 (#1117)
* Fixes for issues 1106 (Graph filters), 1084 (Tree Branch population), and 1116 (Device Template won't set to none)
* Updated changelog\nfor issues #1084, #1106 and #1116
* Cleanup after pull request #1177
Cleanup includes some misqueues in the pull request as well as some
other miscelaneous changes associated with how we have landed on
formatting.
Chagne include:
* No soft tabs on initial tabbing
* ON clause of JOIN should be capitalized and on separate line
* AND/OR should be capitalized in SQL Queries
* Formatting of db_..() functions calls
* Use of sizeof() instead of isset() for array checks
* i18n of Secpass check return values
* Remove trailing whitespace from lines
- ToDo: Remove more having in place of WHERE clauses
* Fix issue #965 - LDAP search error results (#1112)
* Fix issue #965 - LDAP search erorr results
Corrected the error number returned by "Unable to set referrrals" to the correct number.
Added a new error for finding a users DN but not finding the user from the DN.
* Updated changelog for issue #965
* Fix issue 1106 - Graph Filters: template filter have empty value (#1114)
he filter uses index 0 as "All Graphs and Templates" but then searches graph_local for matches against graph_template_id. However, if index 0 is passed as a variable, it is also used in the search. Now when I looked at my database, the following entry existed:
select * from graph_local where graph_template_id = 0;
+------+-------------------+---------+---------------+---------------------+------------+
| id | graph_template_id | host_id | snmp_query_id | snmp_query_graph_id | snmp_index |
+------+-------------------+---------+---------------+---------------------+------------+
| 1837 | 0 | 0 | 0 | 0 | |
+------+-------------------+---------+---------------+---------------------+------------+
Therefore, it displays an index, but can't get a name for the template since there is no template with ID 0.
select * from graph_templates where id = 0;
Empty set (0.00 sec)
Added code to check for a zero value and only search the databsae if a value is present.
* Fix issue 1109 - Opening graph links by unlogged user - validation error (#1115)
* Fix issue 1109 - Opening graph links by unlogged user - validation error
Referer variable needs to be checked to ensure that it doesn't already have parameters applied. If it does, then it needs to use the & rather than ? to join onto the URL.
* Updated changelog for issue #1109
* Fix spelling issue
* Minor updates to pull request 1109.
* Resolving Issue #1104
Different Timestamps in Cacti log when in Developer Debug
* Fix issue #1119 - ss_host_cpu.php pulling back too much data (#1120)
* As per the discussion on #1110 it would appear that ss_host_cpu is pulling back everything from the hrProcessorTable rather than just the processor usage. There appears to be a field named 'hrProcessorFrwID' being defined in the database which is actually the index field not the firmware ID
* Added CHANGELOG for issue #1119
* Fix for issue #1111 - Issue with ldap authenication (#1118)
* Update change log for pull request #1111
* Fixes for issues #1084, #1106 and #1116 (#1117)
* Fixes for issues 1106 (Graph filters), 1084 (Tree Branch population), and 1116 (Device Template won't set to none)
* Updated changelog\nfor issues #1084, #1106 and #1116
* Cleanup after pull request #1177
Cleanup includes some misqueues in the pull request as well as some
other miscelaneous changes associated with how we have landed on
formatting.
Chagne include:
* No soft tabs on initial tabbing
* ON clause of JOIN should be capitalized and on separate line
* AND/OR should be capitalized in SQL Queries
* Formatting of db_..() functions calls
* Use of sizeof() instead of isset() for array checks
* i18n of Secpass check return values
* Remove trailing whitespace from lines
- ToDo: Remove more having in place of WHERE clauses
* Fix issue #965 - LDAP search error results (#1112)
* Fix issue #965 - LDAP search erorr results
Corrected the error number returned by "Unable to set referrrals" to the correct number.
Added a new error for finding a users DN but not finding the user from the DN.
* Updated changelog for issue #965
* Fix issue 1106 - Graph Filters: template filter have empty value (#1114)
he filter uses index 0 as "All Graphs and Templates" but then searches graph_local for matches against graph_template_id. However, if index 0 is passed as a variable, it is also used in the search. Now when I looked at my database, the following entry existed:
select * from graph_local where graph_template_id = 0;
+------+-------------------+---------+---------------+---------------------+------------+
| id | graph_template_id | host_id | snmp_query_id | snmp_query_graph_id | snmp_index |
+------+-------------------+---------+---------------+---------------------+------------+
| 1837 | 0 | 0 | 0 | 0 | |
+------+-------------------+---------+---------------+---------------------+------------+
Therefore, it displays an index, but can't get a name for the template since there is no template with ID 0.
select * from graph_templates where id = 0;
Empty set (0.00 sec)
Added code to check for a zero value and only search the databsae if a value is present.
* Fix issue 1109 - Opening graph links by unlogged user - validation error (#1115)
* Fix issue 1109 - Opening graph links by unlogged user - validation error
Referer variable needs to be checked to ensure that it doesn't already have parameters applied. If it does, then it needs to use the & rather than ? to join onto the URL.
* Updated changelog for issue #1109
* Fix spelling issue
* Minor updates to pull request 1109.
* Resolving Issue #1104
Different Timestamps in Cacti log when in Developer Debug
* Fix issue #1119 - ss_host_cpu.php pulling back too much data (#1120)
* As per the discussion on #1110 it would appear that ss_host_cpu is pulling back everything from the hrProcessorTable rather than just the processor usage. There appears to be a field named 'hrProcessorFrwID' being defined in the database which is actually the index field not the firmware ID
* Added CHANGELOG for issue #1119
* Add CPU Total to 'SNMP - Get Processor Information'
This change adds a virtual index 'Total' to the SNMP info,
showing the average CPU utilization over all cores.
* Changing CPU Total in 'SNMP - Get Processor Information'
Making the total CPU utilization the first element in the array to avoid shifting when CPUs are added.
* Update CHANGELOG
* As per the discussion on #1110 it would appear that ss_host_cpu is pulling back everything from the hrProcessorTable rather than just the processor usage. There appears to be a field named 'hrProcessorFrwID' being defined in the database which is actually the index field not the firmware ID
* Added CHANGELOG for issue #1119
* Fix issue 1109 - Opening graph links by unlogged user - validation error
Referer variable needs to be checked to ensure that it doesn't already have parameters applied. If it does, then it needs to use the & rather than ? to join onto the URL.
* Updated changelog for issue #1109
* Fix spelling issue
he filter uses index 0 as "All Graphs and Templates" but then searches graph_local for matches against graph_template_id. However, if index 0 is passed as a variable, it is also used in the search. Now when I looked at my database, the following entry existed:
select * from graph_local where graph_template_id = 0;
+------+-------------------+---------+---------------+---------------------+------------+
| id | graph_template_id | host_id | snmp_query_id | snmp_query_graph_id | snmp_index |
+------+-------------------+---------+---------------+---------------------+------------+
| 1837 | 0 | 0 | 0 | 0 | |
+------+-------------------+---------+---------------+---------------------+------------+
Therefore, it displays an index, but can't get a name for the template since there is no template with ID 0.
select * from graph_templates where id = 0;
Empty set (0.00 sec)
Added code to check for a zero value and only search the databsae if a value is present.
* Fix issue #965 - LDAP search erorr results
Corrected the error number returned by "Unable to set referrrals" to the correct number.
Added a new error for finding a users DN but not finding the user from the DN.
* Updated changelog for issue #965
Cleanup includes some misqueues in the pull request as well as some
other miscelaneous changes associated with how we have landed on
formatting.
Chagne include:
* No soft tabs on initial tabbing
* ON clause of JOIN should be capitalized and on separate line
* AND/OR should be capitalized in SQL Queries
* Formatting of db_..() functions calls
* Use of sizeof() instead of isset() for array checks
* i18n of Secpass check return values
* Remove trailing whitespace from lines
- ToDo: Remove more having in place of WHERE clauses
- suppress lines breaks breaking graph output if error reporting level is high.
- avoid that remote pollers are requesting graphs from the main poller if the RRDproxy will be used as data storage. In this case remote pollers should directly talk to the proxy.