| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
| |
Fixes:
https://fedorahosted.org/sssd/ticket/2269
sssctl sconfig-check command allows to
call SSSD config file validators on
demand.
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
| |
The timestamp cache tests look into ldb to check the timestamps. This
patch converts the lookups to qualified names to make sure the lookups
actually match.
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
|
| |
The name is converted from whatever we receive on input to the internal
format before processing the data further.
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Adds a utility function sized_output_name() which wraps the output_name()
function and returns the sized_struct structure. This function is used
when formatting the output name for the client, but also when
saving/deleting the memory cache entries.
Its sister function sized_member_name() is very similar, but infers the
domain name from memberuid or ghost attribute.
Because all names internally are used in the same format, the logic to
append domain or format the usename for output in the fill_XXX() family
of functions is much simpler. In general, adding a domain suffix no
longer relies in the domain being a subdomain, but only the dom->fqnames
The parse_member() function was removed because it is no longer
required.
The nss test was amended to store names in the internal fqdn format on
input and checks for either shortnames or qualified names with the right
format created using sss_tc_fqname() on output.
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
|
|
| |
Using a cmocka-based test allows us to initialize the domain using the
common helper functions which in turn allows us to set different
properties with confdb, same as sssd itself does.
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
The sysdb tests now user the qualified name to store users and groups.
To avoid the sysdb interface being tied too tightly to our specific
format, all names are constructed using a function, not hardcoded. Just
swapping the functions that create or parse the names for a different
format should not make the test fail.
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
| |
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
cope with qualified names in sysdb
Many tests use and rely on non-qualified usernames. To avoid huge
commits, we will fix them one-by-one. To avoid test failures in the
branch, disable all the tests first.
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
| |
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
|
|
|
| |
Generates a master key file if it doesn't exist and encrypts secrets
using the master key contained in the file.
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Also adds support for the basic LOCAL provider that stores data
on the local machine.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Start implementing the Secrets Service Reponder core.
This commit implements stratup and basic conenction handling and HTTP
parsing (using the http-parser library).
Signed-off-by: Simo Sorce <simo@redhat.com>
Related:
https://fedorahosted.org/sssd/ticket/2913
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Add helper that uses systemd socket activation if available to accept a
pre-listining socket at startup.
Related:
https://fedorahosted.org/sssd/ticket/2913
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The watchdog uses a kernel timer to issue a signal to the process.
It checks if the ticker is not being reset by the main event loop, which
would indicate that the process got stuck.
At the same time it sets a tevent timer to clear the watchdog ticker, so
that the watchdog handler is kept happy.
If the watchdog detects that the timer event failed to reset the watchdog for
three times in a row then the process is killed.
Normally the monitor will detect the child terminated and will rescheduled it.
Related:
https://fedorahosted.org/sssd/ticket/2921
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Pavel Březina <pbrezina@redhat.com>
|
|
|
|
|
|
|
|
| |
Resolves:
https://fedorahosted.org/sssd/ticket/2028
Reviewed-by: Michal Židek <mzidek@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Resolves:
https://fedorahosted.org/sssd/ticket/2247
Signed-off-by: Lukas Slebodnik <lslebodn@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
| |
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
| |
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
| |
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
|
|
|
| |
So tools (especially sssctl) may be run even when databases where
removed.
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
| |
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
|
| |
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
To be shared across tools and libs.
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
| |
The sysdb initialization was in the sysdb.c module. With adding
initialization of the timestamp cache, this module would become too big
with too many private functions meant only for initialization.
This patch only moves code around, there are no functional changes to
the initialization.
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
|
|
| |
With this plugin winbind can use the same id-mapping as SSSD which makes
it possible to run both together in a consistent way.
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is just a beginning of new responder interface to data provider
and it is just to make the client registration work. It needs further
improvement.
The idea is to take the existing interface and make it work better
with further extensions of data provider. The current interface has
several disadvantages such as it is originally build only for
account requests and doesn't take different set of output parameters.
It also doesn't work well with integration into tevent-made responders.
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Terminology:
* Backend: Implemenation of domain
* Data Provider: interface between backend and responders
* Module: ldap/ipa/ad/... dlopened library that implements dp interface
* Target: id/autofs/sudo/... functionality of module
Benefits over current code:
* data provider is a black box completely separated from backend
* method handlers are just simple tevent requests on backend side
* no need of spy on be_client
* simplified and error proof adding of new responders
* simplified adding of new methods
* reply to D-Bus message is completely handled by DP code
* each target can have several methods defined
* properties can be added on objects
* each method can have output parameters
* modules now support constructor
* improved debugging
* clear memory hierarchy
* ability to chain requests
* type safe private data
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
| |
This patch adds new option 'neg_cache_locals_timeout' into section
of NSS responder. It allows negative caching of local groups and
users. Default value is 0 which means no caching.
Resolves:
https://fedorahosted.org/sssd/ticket/2928
Reviewed-by: Pavel Březina <pbrezina@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When all the dependencies are installed, run the script using systemtap:
You'll see an output such as:
# stap /usr/share/sssd/systemtap/nested_group_perf.stp
^CTime spent in group sssd_be searches: 600
Time spent in sdap_nested_group_send/recv: 65 ms (ratio: 10.83%)
Time spent in zero-level sysdb transactions: 6813 ms (ratio: 1135.50%)
Breakdown of sdap_nested_group req (total: 65 ms)
sdap_nested_group_process req: 64
sdap_nested_group_process_split req: 22
sdap_nested_group_check_cache: 21
sdap_nested_group_sysdb_search_users: 10
sdap_nested_group_sysdb_search_groups: 9
ldap request breakdown of total 60
sdap_nested_group_deref req: 24
sdap_deref_search_send req 24
processing deref results: 0
sdap_nested_group_lookup_user req: 18
sdap_nested_group_lookup_group req: 0
Time spent refreshing unknown members: 18
Breakdown of results processing (total 6813)
Time spent populating nested members: 10
Time spent searching ldb while populating nested members: 5
Time spent saving nested members: 110
Time spent writing to the ldb: 678 ms
Please note that since the script is supposed to be used in scenarios such as
tracing "id" performance, which typically involve multiple group requests.
Therefore, the variables are not zeroed out and you need to interrupt the
script manually with Ctrl+C.
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Run this script using "stap" as root:
sudo stap /path/to/sssd/contrib/systemtap/id_perf.stp
It is not required to restart the script between successive id runs,
the variables are cleared when systemtap detects id had started or finished.
You should see output as this one:
Total run time of id was: 112 ms
Number of zero-level cache transactions: 9
Time spent in level-0 sysdb transactions: 84 ms
Time spent writing to LDB: 80 ms
Number of LDAP searches: 13
Time spent waiting for LDAP: 11 ms
LDAP searches breakdown:
Number of user requests: 1
Time spent in user requests: 15
Number of group requests: 6
Time spent in group requests: 71
Number of initgroups requests: 1
Time spent in initgroups requests: 20
Unaccounted time: 17 ms
sysdb transaction breakdown:
1 hits of transaction sysdb_transaction_commit+0x6b [libsss_util.so]
sdap_save_users+0x2d2 [libsss_ldap_common.so]
sdap_get_users_done+0x186 [libsss_ldap_common.so]
sdap_search_user_process+0x2d9 [libsss_ldap_common.so]
generic_ext_search_handler+0x22f [libsss_ldap_common.so]
sdap_get_and_parse_generic_done+0x6f [libsss_ldap_common.so]
sdap_get_generic_op_finished+0x806 [libsss_ldap_common.so]
sdap_process_message+0x3c4 [libsss_ldap_common.so]
sdap_process_result+0x33a [libsss_ldap_common.so]
sdap_ldap_next_result+0x2f [libsss_
avg:14 min: 14 max: 14 sum: 14
5 hits of transaction sysdb_transaction_commit+0x6b [libsss_util.so]
sdap_nested_done+0x2c8 [libsss_ldap_common.so]
sdap_nested_group_done+0x9b [libsss_ldap_common.so]
0x7f47a6320be4 [libtevent.so.0.9.26+0x4be4]
avg:9 min: 9 max: 11 sum: 49
1 hits of transaction sysdb_transaction_commit+0x6b [libsss_util.so]
sdap_nested_done+0x2c8 [libsss_ldap_common.so]
sdap_nested_group_done+0x9b [libsss_ldap_common.so]
sdap_nested_group_process_done+0x1dc [libsss_ldap_common.so]
sdap_nested_group_single_done+0x112 [libsss_ldap_common.so]
sdap_nested_group_recurse_done+0x95 [libsss_ldap_common.so]
0x7f47a6320be4 [libtevent.so.0.9.26+0x4be4]
avg:11 min: 11 max: 11 sum: 11
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
During performance analysis, the LDAP provider and especially its nested
group code proved to be the place where we spend the most time during
account requests. Therefore, I decorated the LDAP provider with
systemtap probes to be able to observe where the time is spent.
The code allows passing of search properties (base, filter, ...) from
marks to probes. Where applicable, the probes pass on these arguments to
functions and build a human-readable string representation.
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
| |
The caller of the systemtap script would be able to see what kind of
account request sssd received with a string representation, not just the
cryptic hexadecimal number.
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Actually adds marks for sysdb transactions that receive the transaction
nesting level as an argument. The nesting is passed on from probes to
marks along with a human-friendly description.
The transaction commit is decorated with two probes, before and after.
This would allow the caller to distinguish between the time we spend in
the transaction (which might be important, because if a transaction is
active on an ldb context, even the readers are blocked before the
transaction completes) and the time we spend commiting the transaction
(which is important because that's when the disk writes occur)
The probes would be installed into /usr/share/systemtap/tapset on RHEL
and Fedora. This is in line with systemtap's paths which are described
in detail in "man 7 stappaths".
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Adds infrastructure that generatest the probes.h and probes.o from the
dtrace probes.d file. The probes.d file is empty except for the provider
name in this commit, its content will be added with later commits that
actually add some content. The probes.d file is always distributed in
the tarball so that distributions can optionally enable systemtap
support.
The generation is done using the "dtrace" command because the probes.d file
is compatible with the Solaris dtrace format. Please see "man 1 dtrace"
for more information on the dtrace format and the command line tool.
In order to make libtool happy, a fake libtool object is generated. This
hunk was taken from the libvirt code.
The AM_V_GEN macro is used to make the build compatible with the silent
build configuration.
To enable systemtap probing, configure sssd with:
--enable-systemtap
In order to do so, the 'dtrace' command-line utility must be installed.
On Fedora and RHEL, this package is installed as part of the
"systemtap-sdt-devel" package.
You'll also want the 'systemtap' package installed as well as the matching
versions of kernel-devel and kernel-debuginfo on your machine.
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The macros are inspired by very similar macros in libvirt code.
Adds a macro PROBE that can be used by SSSD developers to add systemtap
marks to code. These marks, when coupled with a location in a binary can
be in turn used to call probes. The mark can be called like this:
PROBE(PROBE_NAME, arguments)
This is cleaner than using the SSSD_$(PROBE_NAME) directly as it
directly shows that a probe is being called at that place.
If the systemtap tracing is disabled, they would expand to an empty macro. If
the systemtap tracing is enabled, the systemtap probe will be called.
The overhead of calling the probes is close to zero. As one of the
systemtap developers explained to me:
"""
STAP_PROBE() macros cost apprx. one nop in the executable, so apprx.
no cost at all. The more the merrier. Only when activated by a
stap script do we generally think of it like a microsecond of time.
"""
The probe arguments can be used in the probes to be printed or passed
on to functions. There was an issue in case a string argument was NULL.
This commit adds a helper macro to deal with NULL-strings as if they were
empty ("").
This file would be included by any source file that wants to call the
PROBE() macro.
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|