Differences between current version and predecessor to the previous major change of pam_tally(8).
Other diffs: Previous Revision, Previous Author, or view the Annotated Edit History
Newer page: | version 7 | Last edited on Friday, August 18, 2006 12:14:21 am | by AristotlePagaltzis | |
Older page: | version 4 | Last edited on Thursday, August 17, 2006 3:56:59 am | by TomGreen | Revert |
@@ -1,9 +1,12 @@
-The pam_tally.so PAM module is
intended to denying
further authentication attempts after a given count of failed authentications.
-pam_tally is the maintenance program for pam_tally.so: it can list and reset the accumulated counts
.
+The maintenance program for <tt>
pam_tally.so</tt>. [pam_tally | http://www.baverstock.org.uk/tim/pam/] is a [
PAM]
module intended to deny
further authentication attempts after a given count of failed authentications.
-The
pam_tally project homepage is here:
-http:
//www.baverstock.org.uk
/tim/pam
/
+pam_tally(8) can list and reset the accumulated counts. Note that this only uses a local file (defaults to <tt>
/var
/adm
/faillog<
/tt>) and has no facility to use [LDAP] or similar systems to combine results from several machines (or a cluster).
-Note that this only uses
a local file (defaults to
/var
/adm
/faillog
) and has no facility to use LDAP or similar systems to combine results from several machines
(or a cluster
).
+pam_tally provides
a subset of the functionality of [pam_abl | http:
//www.hexten.net
/pam_abl/], but where pam_tally simply counts failing usernames, pam_abl allows for:
+* counting failing hosts as well as usernames (my logs show the same attacking hostnames trying lots of different usernames rather than the other way round
)
+* configurable time-based failures
(e.g. record a failure if the user or host fails 5 times in an hour
or 10 in
a day
)
+* configurable time-based auto-purging of failure database
-pam_tally provides similar functionality
to pam_abl
(http://www.hexten
.net/pam
_abl/).
+On the other hand,
pam_abl seems
to have 2 issues at the moment:
+* some users
(including TomGreen) report failures not being recorded in database (fixed in current [CVS] from SourceForge)
+* an issue with OpenSSH [where failures don't seem to be recorded |
http://sourceforge
.net/tracker/?group
_id=148927&atid=773100]