IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
Add ctdb_read_nodes_cmd a function that works similarly to
ctdb_read_nodes_file but reads the nodes list from the stdout of a
subprocess instead of a file in the file system.
Signed-off-by: John Mulligan <jmulligan@redhat.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
While here, fix a trivial memory leak (ctdbd will exit anyway if this
function fails).
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Tue Jul 23 12:39:18 UTC 2024 on atb-devel-224
ctdb_control_getnodesfile() calls ctdb_read_nodes(), which returns a
struct ctdb_node_map rather than the old version, so update associated
marshalling. While here modernise a debug message and wrap the
function arguments.
For ctdb_load_nodes_file() to use ctdb_read_nodes(), tweak
convert_node_map_to_list() to also use the modern node map structure.
Remove unused copy of ctdb_read_nodes_file().
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Anoop C S <anoopcs@samba.org>
These functions are intended to be used in ctdbd, the ctdb tool and
fake_ctdbd, replacing the different copies in each place.
ctdb_read_nodes() will replace ctdb_read_nodes_file(). The name
change is intentional - in future the location may be something other
than a simple filename.
The static copies of ctdb_read_nodes_file() and node_map_add() are
slightly sanitised versions of those in tools/ctdb.c, with a call to
ctdb_parse_node_address(). A bit more care is taken in node_map_add()
to avoid undefined behaviour if talloc_realloc() fails.
ctdb_parse_node_address() will replace ctdb_parse_address(). There is
an obvious argument change, since the ctdb context argument was
unused. It can only fail on an invalid node address, so return a
bool. This function might be changed later to allow the input address
string to include an optional port.
Where to put this module isn't entirely clear. It could go in common,
so be part of ctdb-util. However, if it later needs
ctdb-conf (e.g. to allow the node list location to be configurable)
then there would be a direct cyclic dependency. This is configuration
handling, so conf/ seems sane. However, I didn't want to put it into
the ctdb-conf target, since some code might need to parse a nodes list
but not need to parse ctdb.conf.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Anoop C S <anoopcs@samba.org>
The current fake_ctdbd code for reloading the nodes file overruns the
allocation when adding a deleted node at the end. This is a very
unlikely case, but it might as well work.
Check the size of the internal node map when marking a node deleted.
Also, update the code that adds a node to correctly set the deleted
flag when appropriate.
The included test case tests this.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Autobuild-User(master): Günther Deschner <gd@samba.org>
Autobuild-Date(master): Wed Jul 17 00:06:53 UTC 2024 on atb-devel-224
There are no existing tests to exercise node IP address change
detection.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Fails with some compilers with
error: expected ';', ',' or ')' before 'lineptr'
Signed-off-by: Björn Baumbach <bb@sernet.de>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Jo Sutton <josutton@catalyst.net.nz>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Tue Jul 2 23:52:37 UTC 2024 on atb-devel-224
It has been a while since --with-libcephfs option was dropped. Therefore
stop advertising it through waf scripts.
Signed-off-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Tue Jul 2 09:13:20 UTC 2024 on atb-devel-224
CTDB uses a queue to receive requests and send answers. It works
asynchronously using the tevent framework. However there was an issue
that gave priority to the receiving side so, when a request was
processed and the answer posted to the queue, if another incoming
request arrived, it was served before sending the previous answer.
This scenario could repeat for long periods of time if the frequency of
incoming requests was high enough.
Eventually, a small time gap between incoming request gave a chance to
process the pending output queue, sending many answers in a burst.
This patch makes sure that both queues (input and output) are processed
if the event contains the appropriate flag.
Signed-off-by: Xavi Hernandez <xhernandez@redhat.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Mon Jul 1 09:17:43 UTC 2024 on atb-devel-224
We might end up using it elsewhere.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Leave common/conf.[ch] where they are to make this commit
comprehensible.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Guenther Deschner <gd@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
rpc.statd is single-threaded and runs its HA callout synchronously. If
it is too slow then latency accumulates and rpc.statd's backlog grows.
Running a pair of add-client/del-client events with the current code
averages ~0.030s in my test environment. This mean that 1000 clients
reclaiming locks after failover can easily cause 10s of latency. This
could cause rpc.statd to become unresponsive, resulting in a time out
for an rpcinfo-based health check of the status service.
Split the add-client/del-client events out to a standalone
statd_callout executable, written in C, to be used as the HA callout
for rpc.statd. All other functions move to statd_callout_helper.
Now, running a pair of add-client/del-client events in my test
environment averages only ~0.002s. This seems less likely to cause
latency problems.
The standalone statd_callout executable needs to read a configuration
file, which is generated by statd_callout_helper from the "startup"
event. It also needs access to a list of currently assigned public
IPs.
For backward compatibility, during installation a symlink is created
from $CTDB_BASE/statd-callout to the new statd_callout, which is
installed in the helper directory.
Testing this as part of the eventscript unit tests starts to become
even more of a hack than it used to be. However, the dependency on
stubs and the corresponding setup of fake state makes it hard to move
this elsewhere.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Tue Jun 25 04:24:57 UTC 2024 on atb-devel-224
There is a typo here, since there will be no process called "status".
Instead of fixing it, drop this because rpc.statd isn't the focus of
this monitoring check and when systemd is init rpc.statd isn't
restarted with nfs-ganesha. It stays running, so a confusing stack
trace for rpc.statd is always logged.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
If ganesha.nfsd is gone then a node can't provide an NFS service, so
should be marked unhealthy. A later restart may bring it back to
health.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
This one does an rpcinfo check, along with statistics mitigation. It
can be used in combination with the existing 20.nfs_ganesha.check.
The equivalent kernel NFS file only restarts every 10 failures. This
one can be a little more proactive given that false positives are less
likely with the statistics mitigation.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Simplicity is preferred here over absolute correctness. If the
ganesha_stats command exits with an error or times out then no output
is produced so, implicitly, the statistics do not change. Also, the
statistics always change at startup. However, it is likely that the
statistics change when NFS makes progress and do not change when NFS
does not make progress.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
When monitoring an RPC service, the rpcinfo command might time out
even though the service is making progress. In this case, it is just
slow, so counting the timeout as a failure and potentially restarting
the service will not help. The problem is determining if a service is
making progress.
Add a new NFS checks service_stats_command. This command is intended
to run a statistics command. The output is naively compared using
cmp(1). If the output changes then rpcinfo failures are converted to
successes.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Document the new optional argument to specify the namespace to be
associated with RADOS objects in a pool.
Pair-Programmed-With: Anoop C S <anoopcs@samba.org>
Signed-off-by: Günther Deschner <gd@samba.org>
Reviewed-by: Günther Deschner <gd@samba.org>
Reviewed-by: David Disseldorp <ddiss@samba.org>
Autobuild-User(master): Anoop C S <anoopcs@samba.org>
Autobuild-Date(master): Fri Jun 14 07:42:25 UTC 2024 on atb-devel-224
RADOS objects within a pool can be associated to a namespace for
logical separation. librados already provides an API to configure
such a namespace with respect to a context. Make use of it as an
optional argument to the helper binary.
Pair-Programmed-With: Anoop C S <anoopcs@samba.org>
Signed-off-by: Günther Deschner <gd@samba.org>
Reviewed-by: Günther Deschner <gd@samba.org>
Reviewed-by: David Disseldorp <ddiss@samba.org>
While the PID check is worth it in relevant cases, NFS-Ganesha still
might go away after the check. Unfortunately, neither grace command
fails an indicative exit code, so invent one by checking error
messages. This can then be converted to success by the caller.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Thu May 30 12:50:01 UTC 2024 on atb-devel-224
If monitoring has failed because it isn't running, then don't fail
"startipreallocate" or "relaseip" by trying to go into grace.
Don't check this for "takeip". In that case NFS-Ganesha had better be
running.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
No need to grovel around in /proc. ps will happily tell us the
command.
Factor out the actual check into a separate function that can be used
elsewhere.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
Path values do not need to have quotes. The current code fails if
there aren't any.
Instead, implement a 2 stage parser using 2 sed commands. See
comments in the code for details.
Regexps are POSIX basic regular expressions, apart from \<WORD\> (used
to ensure WORD is on word boundaries, and the 'i' flag for case
insensitivity. The latter is supported in FreeBSD sed.
This code successfully parses Path values out of the following
monstrosity:
path = "/foo/bar1;a";
Path = /foo/bar2;
Something = false;
Pseudo = "/foo/bar3x" ; Path = "/foo/bar3; y" ; Access_type = RO;
Pseudo = "/foo/bar4x" ; path=/foo/bar4; Access_type = RO;
Pseudo = "/foo/barNONONO" ; not_Path=/foo/barNONONO; Access_type = RO;
Path = /foo/bar5
Pseudo = "/foo/bar6x Path=foo" ; Path=/foo/bar6; Access_type = RO
This is probably the best that can be done within a shell script.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
Exports may be contained in an include file rather than the top-level
ganesha.conf.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
An IP address is passed to these actions.
Reported-by: Arnab Tah <atah@ddn.com>
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
This simplifies and removes a bad hack. Also, in my test environment,
it also drops the average time take to run an add-client/del-client
pair from ~0.055s to ~0.030s.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
Take advantage of new function find_statd_sm_dir() when clearing the
local system statd state directory, so it uses the correct directory
when running on a non-RH distro.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
For add-client and del-client, statd-callout is called by rpc.statd,
which runs as rpcuser, statd or some other non-root system user. This
means that add-client and del-client can't write in the statd-callout
state directory if it is only writable by root. rpc.statd must be
able to write to its own local system statd state directory, so find
this directory and use it as a reference to set the ownership of
CTDB's statd-callout state directory.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
rpc.statd runs statd-callout as a non-root user, which is currently
hacked around using some sudo logic that fails to work in some
contexts (e.g. in a container).
Use $CTDB_MY_PUBLIC_IPS_CACHE to access the node's currently assigned
public IPs, for add-client/del-client. This avoids connecting to
ctdbd when called from rpc.statd.
Also, use $CTDB_MY_PUBLIC_IPS_CACHE in other places where it makes
sense.
Connections to ctdbd are still made in the "notify" action, but this
is always run as root.
In the test code, set the PNN after public addresses setup so that the
cache of assigned IPs correctly initialised.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
This is called in a couple of places without an argument, so give it a
default.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
This is way more complicated than I would like but, as per the
comment, this is due to complexities in the way public IPs work. The
main consumer will be statd-callout, which will then be able to run as
a non-root user.
Also generate the cache file in test code, whenever the PNN is set.
However, this can cause "ctdb ip" to generate a fake IP layout before
public IPs are setup. So, have the "ctdb ip" stub generate the IP
layout every time it is run to avoid it being stale.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
Add new variables statd_callout_state_dir and statd_callout_queue_dir
- the latter is for files queued by add-client/del-client.
Use $statd_callout_queue_dir to avoid a global cd to the queue
directory near the top of the script.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
All of the other uses of ctdb.tdb are in statd-callout.
New variable statd_callout_db makes it easy to change the database
name in future, perhaps even allowing it to be configurable.
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
Tweak some lines to avoid overflowing 80 columns.
Best viewed with "git show -w".
Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
This avoids a compilation error:
../../ctdb/protocol/protocol_util.c: In function ‘ctdb_connection_list_read’:
../../ctdb/protocol/protocol_util.c:787:9: error: ‘ret’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
787 | return ret;
| ^~~
Signed-off-by: Jo Sutton <josutton@catalyst.net.nz>
Reviewed-by: Martin Schwenke <martin@meltin.net>
