[rt-users] Best Practice RE Mail Server

2011-10-25 Thread Wes Modes
I'm pretty sure this has been hashed and rehashed on this list, but a
google search this morning turned up nothing definitive, so I will ask:

I am reconfiguring a twisted RT installation (3.6, but moving toward
4.0).  We already have a mail server (sendmail) running on another server. 

For incoming and outgoing email I thought of three differrent options:

  * *OPTION A: Re-addressed*
Incoming mail comes into the mail server and is readdressed to RT
server and forwarded
Outgoing mail coming from the RT server goes to the mail server and
is readdressed before going out to the world
Pro: one mail server to admin;  Con: pain to set up,
  * *OPTION B: Redirect*
Incoming mail comes into the mail server and is redirected to the RT
server
Outgoing mail coming from the RT server goes to the mail server and
is redirected to the rest of the world
Pro: one mail server, simpler to set up; Con: less of a setup pain,
but still
  * *OPTION C: Direct*
Incoming mail goes straight to the RT server
Outgoing mail coming from the RT server goes out to the world directly
Pro: Simple to set up;  Con: two mail servers to deal with

What is the best practice (or failing that, Most Common Practice) among
RT administrators?

Wes Modes
University of California,
Santa Cruz

RT Training Sessions (http://bestpractical.com/services/training.html)
*  Washington DC, USA — October 31  November 1, 2011
*  Barcelona, Spain — November 28  29, 2011

Re: [rt-users] Best Practice RE Mail Server

2011-10-25 Thread Wes Modes
Thanks for the suggestion. 

With this however, doesn't sendmail have to be running on the Rt Server
anyway to process all the incoming mail and execute RT scripts? 

For simplicities sake, I'm leaning more and more toward all mail coming
into and going out from the RT server.  What are the disadvantages to
this approach?

Wes

On 10/25/2011 1:24 PM, Francisco Jen Ou wrote:
 Hi,

 My suggestion:

 - incoming email to sendmail and aliased by it to something like
 rt_queue_name@rt_server.domain.com (where rt_server.domain.com resolves
 to RT server's internal net IP).

 - outgoing email sent directly by RT server (don't forget to set
 $SMTPFrom correctly)

 This way, there will be only one server to manage (though you will need
 to setup mailgate on RT server once).



 Em 25-10-2011 16:59, Wes Modes escreveu:
 I'm pretty sure this has been hashed and rehashed on this list, but a
 google search this morning turned up nothing definitive, so I will ask:

 I am reconfiguring a twisted RT installation (3.6, but moving toward
 4.0).  We already have a mail server (sendmail) running on another server. 

 For incoming and outgoing email I thought of three differrent options:

   * *OPTION A: Re-addressed*
 Incoming mail comes into the mail server and is readdressed to RT
 server and forwarded
 Outgoing mail coming from the RT server goes to the mail server and
 is readdressed before going out to the world
 Pro: one mail server to admin;  Con: pain to set up,
   * *OPTION B: Redirect*
 Incoming mail comes into the mail server and is redirected to the RT
 server
 Outgoing mail coming from the RT server goes to the mail server and
 is redirected to the rest of the world
 Pro: one mail server, simpler to set up; Con: less of a setup pain,
 but still
   * *OPTION C: Direct*
 Incoming mail goes straight to the RT server
 Outgoing mail coming from the RT server goes out to the world directly
 Pro: Simple to set up;  Con: two mail servers to deal with

 What is the best practice (or failing that, Most Common Practice) among
 RT administrators?

 Wes Modes
 University of California,
 Santa Cruz


 
 RT Training Sessions (http://bestpractical.com/services/training.html)
 *  Washington DC, USA — October 31  November 1, 2011
 *  Barcelona, Spain — November 28  29, 2011
 
 RT Training Sessions (http://bestpractical.com/services/training.html)
 *  Washington DC, USA  October 31  November 1, 2011
 *  Barcelona, Spain  November 28  29, 2011

RT Training Sessions (http://bestpractical.com/services/training.html)
*  Washington DC, USA  October 31  November 1, 2011
*  Barcelona, Spain  November 28  29, 2011


Re: [rt-users] ExternalAuth help needed

2011-01-13 Thread Wes Modes
problems later.
'attr_match_list'   = ['Name',
'EmailAddress',
],
# The mapping of RT attributes on to LDAP attributes
'attr_map'  =  {   'Name' = 'uid',
'EmailAddress' = 'mail',
'RealName' = 'cn',
'ExternalAuthId' = 'uid',
'Gecos' = 'gecos',
'WorkPhone' =
'telephoneNumber',
}

},
# An example SSO cookie service
'My_SSO_Cookie'  = {  
# # The type of service (db/ldap/cookie)
'type'  =  'cookie',
# The name of the cookie to be used
'name'  =  'loginCookieValue',
# The users table
'u_table'   =  'users',
# The username field in the users table
'u_field'   =  'username',
# The field in the users table that uniquely identifies a user
# and also exists in the cookies table
'u_match_key'   =  'userID',
# The cookies table
'c_table'   =  'login_cookie',
# The field that stores cookie values
'c_field'   =  'loginCookieValue',
# The field in the cookies table that uniquely identifies a user
# and also exists in the users table
'c_match_key'   =  'loginCookieUserID',
# The DB service in this configuration to use to lookup the
cookie information
'db_service_name'   =  'My_MySQL'
}
}

);

1;

Any help would be appreciated.  Thanks.

Wes


On 1/12/2011 4:14 PM, Kevin Falcone wrote:
 On Wed, Jan 12, 2011 at 04:01:08PM -0800, Wes Modes wrote:
  [Wed Jan 12 23:31:22 2011] [error]: AUTH FAILED, Couldn't Load Password 
 Encryption Package.
  Error: Can't locate Crypt/MySQL.pm in @INC (@INC contains: 
 /usr/local/rt/lib
 What are you doing to load that?

 You should send along the other parts of your RT_SiteConfig.pm, it
 appears you've got something 'interesting' running.  Did you tell
 RT-Authen-ExternalAuth to look at LDAP and a mysql database?

 -kevin


Re: [rt-users] ExternalAuth help needed

2011-01-12 Thread Wes Modes
Maybe, but then this isn't what is causing the error, no?

Also, memberOf is one way of implementing groups.  That is a one-to-many
record within the user record.  If you don't have memberOf implemented
within your LDAP, then you can have a group record that contains a list
of userIDs. 

Wes


On 1/12/2011 9:08 AM, Joachim Thuau wrote:

  # Does authentication depend on group membership? What group 
 name?

  'group' =  'staff',

  # What is the attribute for the group object that determines 
 membership?

  'group_attr'=  'ou=group,dc=ucsc,dc=edu',

  

 I believe what you want there is:

  

 'group_attr' = 'memberOf'

  

 As in the name of the attribute that would have the value staff (from 
 your config).

  

 HTH

 Jok



Re: [rt-users] ExternalAuth help needed

2011-01-12 Thread Wes Modes
Kevin, you said the error is clear.  but it seems less than clear to
me.  Especially since I have no problem connecting to ldap with
ldapsearch using the same DN.

Are we stalled out here, then?  If this forum can not offer help, any
suggestions where to turn to for help?

In my experience with other software, LDAP is one of the simplest
integrations possible.  In total one has to configure maybe three, maybe
four things.  The server FQDN, maybe the port, the base DN, and maybe
the root DN.  Voila!  That's it!  LDAP integration. 

I have a difficult time believing that RT is so difficult to integrate
with LDAP, that there is so little step-by-step documentation, and that
the user forums offer so little help.  This has been a surprisingly
difficult process, but I'd still like to be proved completely wrong.

Wes

On 1/11/2011 7:43 AM, Kevin Falcone wrote:
 On Mon, Jan 10, 2011 at 06:03:37PM -0800, Wes Modes wrote:
I am using ExternalAuth to connect RT3.8.8 to LDAP.

Detailed documentation seems to be woefully absent, and I've scoured the 
 web and tried the
dozens of conflicting suggestions, so I'm turning to y'all.

Here's the error I get:

  [Tue Jan 11 01:41:56 2011] [critical]: 
 RT::Authen::ExternalAuth::LDAP::_GetBoundLdapObj
  Can't bind: LDAP_INVALID_DN_SYNTAX 34
  
 (/usr/local/rt/plugins/RT-Authen-ExternalAuth/lib/RT/Authen/ExternalAuth/LDAP.pm:467)
 The error seems clear, something in your username or password isn't
 valid DN syntax according to your server.

 Try connecting using the ldapsearch command line client.

 -kevin

Here's the LDAP section from my RT_Authen-ExternalAuth.pm

  'My_LDAP'   =  {
  ## GENERIC SECTION
  # The type of service (db/ldap/cookie)
  'type'  =  'ldap',
  # The server hosting the service
  'server'=  'dir1.library.ucsc.edu',
  ## SERVICE-SPECIFIC SECTION
  # If you can bind to your LDAP server anonymously you should
  # remove the user and pass config lines, otherwise specify them 
 here:
  #
  # The username RT should use to connect to the LDAP server
  'user'  =  'cn=admin,dc=ucsc,dc=edu',
  # The password RT should use to connect to the LDAP server
  'pass'=  'PASSWORD',
  #
  # The LDAP search base
  'base'  =  'ou=people,dc=ucsc,dc=edu',
  #
  # ALL FILTERS MUST BE VALID LDAP FILTERS ENCASED IN PARENTHESES!
  # YOU **MUST** SPECIFY A filter AND A d_filter!!
  #
  # The filter to use to match RT-Users
  'filter'=  '(objectClass=person)',
  # A catch-all example filter: '(objectClass=*)'
  #
  # The filter that will only match disabled users
  'd_filter'  =  '(objectClass=FooBarBaz)',
  # A catch-none example d_filter: '(objectClass=FooBarBaz)'
  #
  # Should we try to use TLS to encrypt connections?
  'tls'   =  0,
  # SSL Version to provide to Net::SSLeay *if* using SSL
  'ssl_version'   =  3,
  # What other args should I pass to Net::LDAP-new($host,@args)?
  'net_ldap_args' = [version =  3   ],
  # Does authentication depend on group membership? What group 
 name?
  'group' =  'staff',
  # What is the attribute for the group object that determines 
 membership?
  'group_attr'=  'ou=group,dc=ucsc,dc=edu',
  ## RT ATTRIBUTE MATCHING SECTION
  # The list of RT attributes that uniquely identify a user

  # This example shows what you *can* specify.. I recommend 
 reducing this

  # to just the Name and EmailAddress to save encountering 
 problems later.
  'attr_match_list'   = ['Name',
  'EmailAddress',
  ],
  # The mapping of RT attributes on to LDAP attributes
  'attr_map'  =  {   'Name' = 'uid',
  'EmailAddress' = 'mail',
  'RealName' = 'cn',
  'ExternalAuthId' = 'uid',
  'Gecos' = 'gecos',
  'WorkPhone' = 
 'telephoneNumber',
  }

  },

What more do you need to know to help me get this working?

Wes


Re: [rt-users] ExternalAuth help needed

2011-01-12 Thread Wes Modes
I've confirmed that the user rjohnson can indeed login to other
services via LDAP.  However, he can't login to rt.

Looking closer, here's all the rt.log says about the failed login attempt:

[Wed Jan 12 23:21:18 2011] [error]: FAILED LOGIN for rjohnson from
128.114.163.50 (/usr/lib/rt/RT/Interface/Web.pm:424)

Where did the original error I reported go?  Beats me?  Here's what
ldap.log says on the LDAP server:

Jan 12 15:21:18 dir1 slapd2.4[2475]: conn=614436 fd=48 ACCEPT from
IP=128.114.163.13:407
52 (IP=0.0.0.0:389)
Jan 12 15:21:18 dir1 slapd2.4[2475]: conn=614436 op=0 BIND
dn=cn=admin,dc=ucsc,dc=edu
method=128
Jan 12 15:21:18 dir1 slapd2.4[2475]: conn=614436 op=0 BIND
dn=cn=admin,dc=ucsc,dc=edu
mech=SIMPLE ssf=0
Jan 12 15:21:18 dir1 slapd2.4[2475]: conn=614436 op=0 RESULT tag=97
err=0 text=
Jan 12 15:21:18 dir1 slapd2.4[2475]: conn=614436 op=1 SRCH
base=ou=people,dc=ucsc,dc=ed
u scope=2 deref=2 filter=((objectClass=person)(uid=rjohnson))
Jan 12 15:21:18 dir1 slapd2.4[2475]: conn=614436 op=1 SRCH
attr=telephoneNumber cn mail
uid gecos uid
Jan 12 15:21:18 dir1 slapd2.4[2475]: conn=614436 op=1 SEARCH RESULT
tag=101 err=0 nentri
es=1 text=
Jan 12 15:21:18 dir1 slapd2.4[2475]: conn=614436 fd=48 closed
(connection lost)
Jan 12 15:21:19 dir1 slapd2.4[2475]: conn=614435 op=4 UNBIND
Jan 12 15:21:19 dir1 slapd2.4[2475]: conn=614435 fd=39 closed

However, when I login as another user, one that has a local rt login, I
get this in the rt.log:

[Wed Jan 12 23:31:22 2011] [critical]: Search for
(ou=group,dc=ucsc,dc=edu=uid=wmodes,ou=people,dc=ucsc,dc=edu)
failed: LDAP_INVALID_DN_SYNTAX 34

(/usr/local/rt/plugins/RT-Authen-ExternalAuth/lib/RT/Authen/ExternalAuth/LDAP.pm:116)
[Wed Jan 12 23:31:22 2011] [error]: AUTH FAILED, Couldn't Load
Password Encryption Package. Error: Can't locate Crypt/MySQL.pm in
@INC (@INC contains: /usr/local/rt/lib
/usr/local/rt/plugins/RT-Extension-ExtractCustomFieldValues/lib
/usr/local/rt/plugins/RT-Authen-ExternalAuth/lib /usr/lib/rt
/usr/lib64/perl5/site_perl/5.8.8/x86_64-linux-thread-multi
/usr/lib/perl5/site_perl/5.8.8 /usr/lib/perl5/site_perl
/usr/lib64/perl5/vendor_perl/5.8.8/x86_64-linux-thread-multi
/usr/lib/perl5/vendor_perl/5.8.8 /usr/lib/perl5/vendor_perl
/usr/lib64/perl5/5.8.8/x86_64-linux-thread-multi
/usr/lib/perl5/5.8.8 . /etc/httpd) at (eval 4588) line 3, DATA
line 440.

Stack:
  [(eval 4588):3]
 

[/usr/local/rt/plugins/RT-Authen-ExternalAuth/lib/RT/Authen/ExternalAuth.pm:330]
 

[/usr/local/rt/plugins/RT-Authen-ExternalAuth/lib/RT/Authen/ExternalAuth.pm:156]
 

[/usr/local/rt/plugins/RT-Authen-ExternalAuth/html/Callbacks/ExternalAuth/autohandler/Auth:25]
  [/usr/lib/rt/RT/Interface/Web.pm:202]
  [/var/rt/html/autohandler:53]

(/usr/local/rt/plugins/RT-Authen-ExternalAuth/lib/RT/Authen/ExternalAuth/DBI.pm:70)
[Wed Jan 12 23:31:22 2011] [error]: My_MySQL AUTH FAILED The
encryption package you gave me ( Crypt::MySQL ) does not support the
encryption method you specified ( password )

(/usr/local/rt/plugins/RT-Authen-ExternalAuth/lib/RT/Authen/ExternalAuth/DBI.pm:99)


It looks like you are right about the group setting being messed up, but
does that explain the failure of the first login?  There may be multiple
things messed up and prove conclusively that two wrongs do not make a right.

Wes




On 1/12/2011 3:05 PM, Kevin Falcone wrote:
 On Wed, Jan 12, 2011 at 03:00:30PM -0800, Wes Modes wrote:
 Kevin, you said the error is clear.  but it seems less than clear to
 me.  Especially since I have no problem connecting to ldap with
 ldapsearch using the same DN.

 Are we stalled out here, then?  If this forum can not offer help, any
 suggestions where to turn to for help?

 In my experience with other software, LDAP is one of the simplest
 integrations possible.  In total one has to configure maybe three, maybe
 four things.  The server FQDN, maybe the port, the base DN, and maybe
 the root DN.  Voila!  That's it!  LDAP integration. 

 I have a difficult time believing that RT is so difficult to integrate
 with LDAP, that there is so little step-by-step documentation, and that
 the user forums offer so little help.  This has been a surprisingly
 difficult process, but I'd still like to be proved completely wrong.
 Wes, plenty of folks have this working and I've set it up more times
 than I can count.  You've cherry picked an error that tells me that
 your LDAP server is rejecting the connection attempt.  What do your
 LDAP logs say?

 As someone observed, your group settings look interestingly wrong, but
 since I'm staring at one log line it's kind of hard to tell if this is
 the initial bind or a later bind failure.

 -kevin

 On 1/11/2011 7:43 AM, Kevin Falcone wrote:
 On Mon, Jan 10, 2011 at 06:03:37PM -0800, Wes Modes wrote:
I am using

Re: [rt-users] ExternalAuth help needed

2011-01-11 Thread Wes Modes
The machine name is specified elsewhere, but that is what the root DN
looks like.  Now whether that is the right format for that variable
value or not, I don't know. 

W.

On 1/11/2011 5:49 AM, Josh Narins wrote:

 I have fiddled only a little with LDAP.

  

 The error message sounds like it isn't recognizing something as a DN.

  

 To me, your username doesn't look quite right. Is there really an LDAP
 server at ucsc.edu? Shouldn't it be more like DC=ldap1,DC=ucsc,DC=edu,
 to specify the machine name?

  

 I'm not even very good with windows, so, I could be way off, here.

  



 //

 *Josh Narins*

 Director of Application Development
 SeniorBridge
 845 Third Ave
 7th Floor
 New York, NY 10022
 Tel: (212) 994-6194
 Fax: (212) 994-4260
 Mobile: (917) 488-6248
 jnar...@seniorbridge.com
 seniorbridge.com http://www.seniorbridge.com/

 SeniorBridge



 
 *SeniorBridge Statement of Confidentiality:* The contents of this
 email message are intended for the exclusive use of the addressee(s)
 and may contain confidential or privileged information. Any
 dissemination, distribution or copying of this email by an unintended
 or mistaken recipient is strictly prohibited. In said event, kindly
 reply to the sender and destroy all entries of this message and any
 attachments from your system. Thank you.

 *From:*rt-users-boun...@lists.bestpractical.com
 [mailto:rt-users-boun...@lists.bestpractical.com] *On Behalf Of *Wes Modes
 *Sent:* Monday, January 10, 2011 9:04 PM
 *To:* RT Users
 *Subject:* [rt-users] ExternalAuth help needed

  

 I am using ExternalAuth to connect RT3.8.8 to LDAP.

 Detailed documentation seems to be woefully absent, and I've scoured
 the web and tried the dozens of conflicting suggestions, so I'm
 turning to y'all.

 Here's the error I get:

 [Tue Jan 11 01:41:56 2011] [critical]:
 RT::Authen::ExternalAuth::LDAP::_GetBoundLdapObj Can't bind:
 LDAP_INVALID_DN_SYNTAX 34
 (/usr/local/rt/plugins/RT-Authen-ExternalAuth/lib/RT/Authen/ExternalAuth/LDAP.pm:467)


 Here's the LDAP section from my RT_Authen-ExternalAuth.pm

 'My_LDAP'   =  {  
 ## GENERIC SECTION
 # The type of service (db/ldap/cookie)
 'type'  =  'ldap',
 # The server hosting the service
 'server'=  'dir1.library.ucsc.edu',
 ## SERVICE-SPECIFIC SECTION
 # If you can bind to your LDAP server anonymously you should
 # remove the user and pass config lines, otherwise specify
 them here:
 #
 # The username RT should use to connect to the LDAP server
 'user'  =  'cn=admin,dc=ucsc,dc=edu',
 # The password RT should use to connect to the LDAP server
 'pass'=  'PASSWORD',
 #
 # The LDAP search base
 'base'  =  'ou=people,dc=ucsc,dc=edu',
 #
 # ALL FILTERS MUST BE VALID LDAP FILTERS ENCASED IN PARENTHESES!
 # YOU **MUST** SPECIFY A filter AND A d_filter!!
 #
 # The filter to use to match RT-Users
 'filter'=  '(objectClass=person)',
 # A catch-all example filter: '(objectClass=*)'
 #
 # The filter that will only match disabled users
 'd_filter'  =  '(objectClass=FooBarBaz)',
 # A catch-none example d_filter: '(objectClass=FooBarBaz)'
 #
 # Should we try to use TLS to encrypt connections?
 'tls'   =  0,
 # SSL Version to provide to Net::SSLeay *if* using SSL
 'ssl_version'   =  3,
 # What other args should I pass to Net::LDAP-new($host,@args)?
 'net_ldap_args' = [version =  3   ],
 # Does authentication depend on group membership? What group name?
 'group' =  'staff',
 # What is the attribute for the group object that determines
 membership?
 'group_attr'=  'ou=group,dc=ucsc,dc=edu',
 ## RT ATTRIBUTE MATCHING SECTION
 # The list of RT attributes that uniquely identify a user

 # This example shows what you *can* specify.. I recommend
 reducing this

 # to just the Name and EmailAddress to save encountering
 problems later.
 'attr_match_list'   = ['Name',
 'EmailAddress',
 ],
 # The mapping of RT attributes on to LDAP attributes
 'attr_map'  =  {   'Name' = 'uid',
 'EmailAddress' = 'mail',
 'RealName' = 'cn',
 'ExternalAuthId' = 'uid',
 'Gecos' = 'gecos',
 'WorkPhone

[rt-users] ExternalAuth help needed

2011-01-10 Thread Wes Modes
I am using ExternalAuth to connect RT3.8.8 to LDAP.

Detailed documentation seems to be woefully absent, and I've scoured the
web and tried the dozens of conflicting suggestions, so I'm turning to
y'all.

Here's the error I get:

[Tue Jan 11 01:41:56 2011] [critical]:
RT::Authen::ExternalAuth::LDAP::_GetBoundLdapObj Can't bind:
LDAP_INVALID_DN_SYNTAX 34

(/usr/local/rt/plugins/RT-Authen-ExternalAuth/lib/RT/Authen/ExternalAuth/LDAP.pm:467)


Here's the LDAP section from my RT_Authen-ExternalAuth.pm

'My_LDAP'   =  {  
## GENERIC SECTION
# The type of service (db/ldap/cookie)
'type'  =  'ldap',
# The server hosting the service
'server'=  'dir1.library.ucsc.edu',
## SERVICE-SPECIFIC SECTION
# If you can bind to your LDAP server anonymously you should
# remove the user and pass config lines, otherwise specify
them here:
#
# The username RT should use to connect to the LDAP server
'user'  =  'cn=admin,dc=ucsc,dc=edu',
# The password RT should use to connect to the LDAP server
'pass'=  'PASSWORD',
#
# The LDAP search base
'base'  =  'ou=people,dc=ucsc,dc=edu',
#
# ALL FILTERS MUST BE VALID LDAP FILTERS ENCASED IN PARENTHESES!
# YOU **MUST** SPECIFY A filter AND A d_filter!!
#
# The filter to use to match RT-Users
'filter'=  '(objectClass=person)',
# A catch-all example filter: '(objectClass=*)'
#
# The filter that will only match disabled users
'd_filter'  =  '(objectClass=FooBarBaz)',
# A catch-none example d_filter: '(objectClass=FooBarBaz)'
#
# Should we try to use TLS to encrypt connections?
'tls'   =  0,
# SSL Version to provide to Net::SSLeay *if* using SSL
'ssl_version'   =  3,
# What other args should I pass to Net::LDAP-new($host,@args)?
'net_ldap_args' = [version =  3   ],
# Does authentication depend on group membership? What group
name?
'group' =  'staff',
# What is the attribute for the group object that determines
membership?
'group_attr'=  'ou=group,dc=ucsc,dc=edu',
## RT ATTRIBUTE MATCHING SECTION
# The list of RT attributes that uniquely identify a user

# This example shows what you *can* specify.. I recommend
reducing this

# to just the Name and EmailAddress to save encountering
problems later.
'attr_match_list'   = ['Name',
'EmailAddress',
],
# The mapping of RT attributes on to LDAP attributes
'attr_map'  =  {   'Name' = 'uid',
'EmailAddress' = 'mail',
'RealName' = 'cn',
'ExternalAuthId' = 'uid',
'Gecos' = 'gecos',
'WorkPhone' =
'telephoneNumber',
}

},


What more do you need to know to help me get this working?

Wes



Re: [rt-users] Why I am recommending 3.6 over 3.8 to my boss

2010-11-05 Thread Wes Modes
Agreed.  One sysadmin managing a score of mission-critical servers and a
half dozen projects does not allow much time for one-offs and special
cases.  Over my 25 years of sysadmin experience, I've learned that the
most efficient thing I can do as a sysadmin is to allow the package
management system to do much of my work for me.

There are legacy systems I inherited with their spaghetti installations
of all special-case software and manual hack builds and their touchy
interdependencies that I am still afraid to do much more than basic
security updates of the OS.

Wes

On 11/5/2010 5:11 AM, Vick Khera wrote:
 On Nov 5, 2010, at 5:26 AM, Robert Grasso wrote:

 This is my own opinion : as you increase your Unix/Linux/RedHat skills, you 
 will feel less concerned by such issues.
 As you increase the number of systems you need to manage, you will feel more 
 concerned by such issues.

 A good package manager to manage all of your software is essential to 
 configuration management on a large scale.  We even go so far as to make 
 internal packages of our own software to deploy to the servers -- nothing is 
 manually done, except for the one-off office server which does the file/mail 
 serving.

 As you note later in your message, you have to manually go in and fix up 
 things when you upgrade other parts of your system.  This is the job of your 
 package manager.  It does not scale to do this by hand.


[rt-users] Why I am recommending 3.6 over 3.8 to my boss

2010-11-04 Thread Wes Modes
Dear Boss:

I strongly recommend going with the 3.6 version of RT.  The install takes a few 
minutes, and it otherwise meets all the requirements of our project.  Migration 
of old queues is simple.  There is cost savings in the near and long-term.  

There is no rpm of RT3.8 that works for RHEL (32 or 64 bit) and none seem to be 
forthcoming.  Someday perhaps someone will put one together, but it doesn't 
look like anytime soon.  

I CAN do a manual install of RT3.8 using the Best Practical install scripts.  
It is not terribly hard.  However, the long-term costs of this are large.  The 
install scripts put all the binaries, configuration files, and libraries in the 
wrong places for RHEL/CentOS, and working outside the package manager means 
files could be clobbered at any time.  On the other hand, the rpms for RT3.6 
use the package manager and put all the config files in /etc, all the perl 
modules in the perl modules dir, and the various tools in /usr/bin and 
/usr/sbin.  The non-standard install using the scripts creates recurring costs 
in the future as the system is significantly more difficult to update and 
harder to maintain, like by a factor of 50 (five minutes compared to 4 hours).

Additionally, the cost of migration of old content from 3.6 to 3.8 is unknown.

Again, I will install either RT3.6 or RT3.8 but I need you to understand
and acknowledge the costs of the choice.

Wes


Thanks to Gary Greene for the info about his latest centos rpm build.



Re: [rt-users] Frustrating attempts to install RT3.8 from RPM

2010-11-03 Thread Wes Modes
Paul, sounds like you aren't a long term fan of Fedora, RHEL, or CentOS,
so I'm guessing yum feels like an inconvenience to you, especially when
it seems to be getting in the way of your desired install.

I've been a sysadmin for 20 years and I've never been a fan of the make
'n' break style of system administration.  There is no way I could
manage a score of machines, many with subtly different hardware, if I
had to build every package the old way.  As it is, I can spend a few
hours monthly updating the OS and all installed software on all of our
machines, with a simple yum -y update

In my opinion, package managers like apt-get and yum are some of the
best things to happen to OS in a very long time.  Having installs
tracked and managed by package managers keeps complicated OSs and their
installed software up-to-date, eases system administration (especially
as the server to sysadmin ratio increases), increases scalability,
increases sysadmin efficiency, and creates standards for software
manufacturers. 

If as a conservative sysadmin you prefer to operate well-back from the
bleeding edge anyway, the small trade-off in control is a small price to
pay.

It is hardly the package manager's fault if a software manufacturer such
as Best Practical and its user community fail to create a package for
the latest software.  Compare that to software whose RPMs are kept
relatively up-to-date. 

Wes

On 11/2/2010 3:49 PM, Paul wrote:
 On 11/02/2010 02:19 PM, Wes Modes wrote:
 Hello, I have been struggling with attempts to install RT3.8 via RPMs.  

 I know it is perfectly possible to install RT3.8 using the BP install
 scripts and docs, but I'd prefer to do it through yum for system
 sustainability, ease of updates and upgrades, etc.
 ...
 If I can't resolve this, I will just forget about RT3.8 and stick with
 RT3.6 of which there is a well-behaved RPM already in the EPEL repo.

 Wes

 I'm currently going through a RT move from freebsd to rhel5 (long story,
 would rather stay with freebsd but don't have a choice here) and have
 found all kinds of annoying difficulties with yum (or, rather, the
 packages available.) When I realized that I was trying to stick with yum
 for ease of upgrades when yum was preventing me from easily keeping up
 to date, life got a lot easier.

 In the end I just let cpan install what it could and used yum for the
 things that gave me trouble in cpan. Using RT's configure and make
 targets is a lot easier and much more maintainable than having to roll
 my own rpm just to do it the yum way.

 Being stuck with an old version of the software in the name of easy
 upgrades didn't make sense to me.

 Cheers,
 Paul


Re: [rt-users] Frustrating attempts to install RT3.8 from RPM

2010-11-03 Thread Wes Modes
I presume that is CentOS5.  That would make me very happy as CentOS RPMs
should work for RHEL.

One thing I adore about well-built packages is that things are placed in
the right location for the OS.  For instance, the RT3 rpms put all the
config files in /etc, all the perl modules in the perl modules dir, and
the various tools in /usr/bin and /usr/sbin.

Is yours built that way, or does it keep to the Best Practical distro
locations?

i guess this means that no one has a solution to the problem I observed
with the rpm bundle I did find, ya?

Wes


On 11/3/2010 11:52 AM, Gary Greene wrote:
 Agreed. This is why I spent a week with cpan2rpm and built packages for both
 openSuSE (which we're transitioning to) and CentOS.


 On 3/11/10 11:21 AM, Wes Modes wmo...@ucsc.edu wrote:

 Paul, sounds like you aren't a long term fan of Fedora, RHEL, or CentOS,
 so I'm guessing yum feels like an inconvenience to you, especially when
 it seems to be getting in the way of your desired install.

 I've been a sysadmin for 20 years and I've never been a fan of the make
 'n' break style of system administration.  There is no way I could
 manage a score of machines, many with subtly different hardware, if I
 had to build every package the old way.  As it is, I can spend a few
 hours monthly updating the OS and all installed software on all of our
 machines, with a simple yum -y update

 In my opinion, package managers like apt-get and yum are some of the
 best things to happen to OS in a very long time.  Having installs
 tracked and managed by package managers keeps complicated OSs and their
 installed software up-to-date, eases system administration (especially
 as the server to sysadmin ratio increases), increases scalability,
 increases sysadmin efficiency, and creates standards for software
 manufacturers. 

 If as a conservative sysadmin you prefer to operate well-back from the
 bleeding edge anyway, the small trade-off in control is a small price to
 pay.

 It is hardly the package manager's fault if a software manufacturer such
 as Best Practical and its user community fail to create a package for
 the latest software.  Compare that to software whose RPMs are kept
 relatively up-to-date.

 Wes

 On 11/2/2010 3:49 PM, Paul wrote:
 On 11/02/2010 02:19 PM, Wes Modes wrote:
 Hello, I have been struggling with attempts to install RT3.8 via RPMs.

 I know it is perfectly possible to install RT3.8 using the BP install
 scripts and docs, but I'd prefer to do it through yum for system
 sustainability, ease of updates and upgrades, etc.
 ...
 If I can't resolve this, I will just forget about RT3.8 and stick with
 RT3.6 of which there is a well-behaved RPM already in the EPEL repo.

 Wes

 I'm currently going through a RT move from freebsd to rhel5 (long story,
 would rather stay with freebsd but don't have a choice here) and have
 found all kinds of annoying difficulties with yum (or, rather, the
 packages available.) When I realized that I was trying to stick with yum
 for ease of upgrades when yum was preventing me from easily keeping up
 to date, life got a lot easier.

 In the end I just let cpan install what it could and used yum for the
 things that gave me trouble in cpan. Using RT's configure and make
 targets is a lot easier and much more maintainable than having to roll
 my own rpm just to do it the yum way.

 Being stuck with an old version of the software in the name of easy
 upgrades didn't make sense to me.

 Cheers,
 Paul


Re: [rt-users] Frustrating attempts to install RT3.8 from RPM

2010-11-03 Thread Wes Modes
That is nice to see that you made a well-crafted rpm that you can be
proud of.   I wonder what would happen if a later version of RT3 became
available via EPEL.  Would it nicely replace the files (maybe moving
stuff to rpmsave's)  or would all hell break loose?

What RT3 version is your centos rpm build?

When and where would your centos rpm be available to play with?

W.

On 11/3/2010 4:45 PM, Gary Greene wrote:
 The CentOS ones follow the RH way of directory layout, with the caveat that
 I chose to put the other modules that normally get pulled in via cpan in the
 perl5 site_lib hierarchy to assure that a rouge update from rpmforge or
 upstream CentOS would be able to be installed without odd file conflicts.

 The SuSE ones I did slightly differently as I think having the main RT stuff
 strewn around /usr a little odd. The CPAN stuff is in the perl5 site_lib
 hierarchy as before, but the main HTML/Mason templates/RT only specific
 modules/plugins stuff are in /srv/www/htdocs/rt. Configuration stuff is in
 /etc/rt and the plugin configuration directory is /etc/rt/local/...

 If I were to do over the CentOS ones, I'd likely do the same as I did with
 SuSE.

 On 3/11/10 4:36 PM, Wes Modes wmo...@ucsc.edu wrote:

 I presume that is CentOS5.  That would make me very happy as CentOS RPMs
 should work for RHEL.

 One thing I adore about well-built packages is that things are placed in
 the right location for the OS.  For instance, the RT3 rpms put all the
 config files in /etc, all the perl modules in the perl modules dir, and
 the various tools in /usr/bin and /usr/sbin.

 Is yours built that way, or does it keep to the Best Practical distro
 locations?

 i guess this means that no one has a solution to the problem I observed
 with the rpm bundle I did find, ya?

 Wes


 On 11/3/2010 11:52 AM, Gary Greene wrote:
 Agreed. This is why I spent a week with cpan2rpm and built packages for both
 openSuSE (which we're transitioning to) and CentOS.


 On 3/11/10 11:21 AM, Wes Modes wmo...@ucsc.edu wrote:

 Paul, sounds like you aren't a long term fan of Fedora, RHEL, or CentOS,
 so I'm guessing yum feels like an inconvenience to you, especially when
 it seems to be getting in the way of your desired install.

 I've been a sysadmin for 20 years and I've never been a fan of the make
 'n' break style of system administration.  There is no way I could
 manage a score of machines, many with subtly different hardware, if I
 had to build every package the old way.  As it is, I can spend a few
 hours monthly updating the OS and all installed software on all of our
 machines, with a simple yum -y update

 In my opinion, package managers like apt-get and yum are some of the
 best things to happen to OS in a very long time.  Having installs
 tracked and managed by package managers keeps complicated OSs and their
 installed software up-to-date, eases system administration (especially
 as the server to sysadmin ratio increases), increases scalability,
 increases sysadmin efficiency, and creates standards for software
 manufacturers. 

 If as a conservative sysadmin you prefer to operate well-back from the
 bleeding edge anyway, the small trade-off in control is a small price to
 pay.

 It is hardly the package manager's fault if a software manufacturer such
 as Best Practical and its user community fail to create a package for
 the latest software.  Compare that to software whose RPMs are kept
 relatively up-to-date.

 Wes

 On 11/2/2010 3:49 PM, Paul wrote:
 On 11/02/2010 02:19 PM, Wes Modes wrote:
 Hello, I have been struggling with attempts to install RT3.8 via RPMs.

 I know it is perfectly possible to install RT3.8 using the BP install
 scripts and docs, but I'd prefer to do it through yum for system
 sustainability, ease of updates and upgrades, etc.
 ...
 If I can't resolve this, I will just forget about RT3.8 and stick with
 RT3.6 of which there is a well-behaved RPM already in the EPEL repo.

 Wes

 I'm currently going through a RT move from freebsd to rhel5 (long story,
 would rather stay with freebsd but don't have a choice here) and have
 found all kinds of annoying difficulties with yum (or, rather, the
 packages available.) When I realized that I was trying to stick with yum
 for ease of upgrades when yum was preventing me from easily keeping up
 to date, life got a lot easier.

 In the end I just let cpan install what it could and used yum for the
 things that gave me trouble in cpan. Using RT's configure and make
 targets is a lot easier and much more maintainable than having to roll
 my own rpm just to do it the yum way.

 Being stuck with an old version of the software in the name of easy
 upgrades didn't make sense to me.

 Cheers,
 Paul


[rt-users] Frustrating attempts to install RT3.8 from RPM

2010-11-02 Thread Wes Modes
Hello, I have been struggling with attempts to install RT3.8 via RPMs.  

I know it is perfectly possible to install RT3.8 using the BP install
scripts and docs, but I'd prefer to do it through yum for system
sustainability, ease of updates and upgrades, etc.

These instructions show how to set up a local repo and install RT from a
bundle, but for version 3.6. 

*Installing RT 3.6.6 on Redhat Enterprise 5.x (using yum to install
from a bundle)*
http://wiki.bestpractical.com/view/Rhel5InstallGuide

(keep this link, because it is hard to find and all of the sometimes
contradictory RT docs look the same)

However there is a similar bundle for 3.8.7, so maybe that would work.

According to the install doc, we install a host of services first:

[r...@testbench1]# yum -y update

[r...@testbench1]# yum -y install httpd

[r...@testbench1]# yum -y install mysql mysql-server sendmail-cf

Start these services:

[r...@testbench1]# service mysqld start
Starting MySQL:[  OK  ]
[r...@testbench1]# service httpd start
Starting httpd:[  OK  ]
[r...@testbench1]# chkconfig httpd on
[r...@testbench1]# chkconfig mysqld on

The docs call for downloading this bundle: 

http://www.jwhite3.com/files/rt/rt-3.6.6-bundle.tar.gz

but we are going to be downloading the 3.8.7 bundle

[r...@testbench1]# cd
[r...@testbench1]# pwd
/root
[r...@testbench1]# mkdir rt3
[r...@testbench1]# cd rt3
[r...@testbench1]# wget http://www.jwhite3.com/files/rt/rt_3.8.7_bundle.zip
--2010-10-29 16:18:39--  http://www.jwhite3.com/files/rt/rt_3.8.7_bundle.zip
Resolving www.jwhite3.com http://www.jwhite3.com... 97.74.144.177
Connecting to www.jwhite3.com|97.74.144.177|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 38577186 (37M) [application/zip]
Saving to: `rt_3.8.7_bundle.zip'

100%[] 38,577,186  4.87M/s   in 
7.8s

2010-10-29 16:18:47 (4.72 MB/s) - `rt_3.8.7_bundle.zip' saved 
[38577186/38577186]

Unpack:

[r...@testbench1]# unzip rt_3.8.7_bundle.zip  
Archive:  rt_3.8.7_bundle.zip
  inflating: install.sh  
  inflating: Modules.tar.gz  
  inflating: rt-3.8.7.tar.gz 
  inflating: rt.repo 
  inflating: rt_repo.tar.gz 

set up yum repo file:

[r...@testbench1]# ls
install.sh  rt_3.8.7_bundle.zip  rt.repo
Modules.tar.gz  rt-3.8.7.tar.gz  rt_repo.tar.gz

[r...@testbench1]# cp rt.repo  /etc/yum.repos.d/

[r...@testbench1]# vi /etc/yum.repos.d/rt.repo

[rt-387-local]
name=Request Tracker - $basearch
baseurl=file://opt/rt_repo/$basearch/
enabled=1
gpgcheck=0

[rt-387-noarch-local]
name=Request Tracker - noarch
baseurl=file://opt/rt_repo/noarch/
enabled=1
gpgcheck=0

Unpack the distro part and move it over to /opt where the yum file
expected it:

[r...@testbench1]# tar xfz rt_repo.tar.gz

[r...@testbench1]# mv rt_repo /opt
[r...@testbench1]# ls /opt/rt_repo/
i386  noarch  x86_64

Okay, let's see if that works:

[r...@testbench1]# yum clean all
Loaded plugins: rhnplugin, security
Cleaning up Everything

[r...@testbench1]# yum list rt3
Loaded plugins: rhnplugin, security
adobe-linux-i386   |  951 B 
00:00 
adobe-linux-i386/primary   |  12 kB 
00:00 
adobe-linux-i386
18/18
rhel-i386-server-5 | 1.4 kB 
00:00 
rhel-i386-server-5/primary | 3.0 MB 
00:00 
rhel-i386-server-5  
7696/7696
rhel-i386-server-vt-5  | 1.4 kB 
00:00 
rhel-i386-server-vt-5/primary  |  41 kB 
00:00 
rhel-i386-server-vt-5 
198/198
rhn-tools-rhel-i386-server-5   | 1.3 kB 
00:00 
rhn-tools-rhel-i386-server-5/primary   |  38 kB 
00:00 
rhn-tools-rhel-i386-server-5  
457/457
file://opt/rt_repo/i386/repodata/repomd.xml: 
http://opt/rt_repo/i386/repodata/repomd.xml: [Errno 5] OSError: [Errno 2] No 
such file or directory: '/rt_repo/i386/repodata/repomd.xml'
Trying other mirror.
Error: Cannot retrieve repository metadata (repomd.xml) for repository: 
rt-387-local. Please verify its path and try again

No clue what this means.  I checked the yum locations.  I checked the
xml metadata.  Couldn't see where this bad path was coming from.

Any suggestions for resolving this?

If I can't resolve this, I will just forget about RT3.8 and stick with
RT3.6 of which there is a well-behaved RPM already in the EPEL repo.

Wes