knet needs buffers to be KNET_MAX_PACKET_SIZE or messages will
get lost or corrupted.
UDPU packets shouldn't be that big so I introduced UDP_FRAME_SIZE_MAX
for that transport.
Signed-off-by: Christine Caulfield <ccaulfie@redhat.com>
segv should be handled by corosync, libqb is not the
place to be handling emergency signals.
This currently requires the head of libqb git tree to
generate a blackbox & coredump in the event of a segfault,
but it's better than the write() spin that currently happens.
Signed-off-by: Christine Caulfield <ccaulfie@redhat.com>
Reviewed-by: Jan Friesse <jfriesse@redhat.com>
LibQB adds default "*" syslog filter so we have to set syslog_priority
as low as possible so filters applied later in
_logsys_config_apply_per_file takes effect.
Signed-off-by: Jan Friesse <jfriesse@redhat.com>
after some feedback on github, people prefers to have the option
to support up to 64K node_id's.
libknet added knet_node_id_t to mask the size and type, currently
set to uint16_t.
Signed-off-by: Fabio M. Di Nitto <fdinitto@redhat.com>
Corosync layers don't need to know the knet MTU size - this way
corosync fragments buffers only when they get larger than the
KNET buffer size (64K) and knet fragments below that based on
the actual MTU and transport considerations.
It is also now possible to configure knet to use UDP or SCTP
transports in corosync.conf. This is currently done per-link
so if you have more than 1 link you need several interface{}
stanzas inside totem{} to make it use other than the default
of UDP. if it's useful I might add the option of a global
default.
Signed-off-by: Christine Caulfield <ccaulfie@redhat.com>
I've also put an assert in totemknet_member_add() to check
for invalid nodeids. Later on we need to fix the rest of the
corosync code to only use 8bit nodeids (or force people to use
UDPU if they want large nodeids).
Signed-off-by: Christine Caulfield <ccaulfie@redhat.com>
Commit 8d8d4a936a introduced the
configuration parameter resources.watchdog_device. This commit
introduces the resources section and watchdog_device parameter in
corosync.conf.5.
Signed-off-by: Adrian Vondendriesch <adrian.vondendriesch@credativ.de>
Reviewed-by: Jan Friesse <jfriesse@redhat.com>
corosynclib-devel doesn't need to have a dependency on corosync package.
It's expected that libraries are still working properly (e.g. indicating
errors to their users) when there's no corosync process around in that
moment, and from this perspective it doesn't matter whether it is
installed at all for some purposes, especially having linkage with them
in mind.
Note that the inverse dependency, main corosync package on corosynclib,
is already there (not strictly needed, likely just to enforce package
version match -- otherwise RPM's dependency generator will handle this
on its own using SONAMEs -- hence the comments to that effect are also
added), so breaking this symmetry:
- is supposed to be harmless modulo cases that should be fixed to
express explicit dependency on corosync's runtime anyway
(but only for runtime, i.e., Requires as opposed to BuildRequires)
- will effectively enable more lightweight get-build-deps-and-build
process for programs linking with corosynclibs (e.g. pacemaker),
as corosync package won't need to be installed needlessly
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
Reviewed-by: Jan Friesse <jfriesse@redhat.com>
There are somwe spell errors in qdevice-net-algo-test.c,
qnetd-algo-2nodelms.c, qnetd-algo-test.c qnetd-algo-ffsplit.c
Signed-off-by: Bin Liu <bliu@suse.com>
Reviewed-by: Jan Friesse <jfriesse@redhat.com>
Running 'doxygen -u Doxyfile.in' in the source root produces the
following results:
- SYMBOL_CACHE_SIZE at line 301 has become obsolete. This tag
has been removed.
- SHOW_DIRECTORIES at line 507 has become obsolete. This tag
has been removed.
- HTML_ALIGN_MEMBERS at line 881 has become obsolete. This tag
has been removed.
- USE_INLINE_TREES at line 1067 has become obsolete. This tag
has been removed.
- XML_SCHEMA at line 1311 has become obsolete. This tag has been
removed.
- XML_DTD at line 1317 has become obsolete. This tag has been
removed.
Signed-off-by: Richard B Winters <rik@mmogp.com>
Reviewed-by: Jan Friesse <jfriesse@redhat.com>
In function mcast_msg of totempg.c, line 923, there is a memcpy call in
"else" branch, and also another memcpy out of the "else" branch, while
the two calls have the same parameters. It is possibleto remove the memcpy
in "else" branch.
Signed-off-by: Bin Liu <bliu@suse.com>
Reviewed-by: Jan Friesse <jfriesse@redhat.com>
totempg_groups_join() is called by sync_init().
sync_init() judge that totempg_groups_join() failed if return code of
totempg_groups_join() is -1.
Therefore, the return code should return in -1 when
totempg_groups_join() fails.
Signed-off-by: Takeshi MIZUTA <miz.take4@gmail.com>
Reviewed-by: Jan Friesse <jfriesse@redhat.com>
vqsim is a small program that allows node up/down/split/join
operations to be simulated without the use of an actual cluster.
Signed-off-by: Christine Caulfield <ccaulfie@redhat.com>
This is a big update that removes RRP & MRP from the codebase
and makes knet the default transport for corosync. UDP & UDPU
are still (currently) supported but are deprecated. Also crypto
and mutiple interfaces are only supported over knet.
To compile this codebase you will need to install libknet from
https://github.com/fabbione/kronosnet
The corosync.conf(5) man page has been updated with info on the new
options. Older config files should still work but many options
have changed because of the knet implementation so configs should
be checked carefully. In particular any cluster using using RRP
over UDP or UDPU will not start as RRP is no longer present. If you
need multiple interface support then you should be using the knet transport.
Knet brings many benefits to the corosync codebase, it provides support
for more interfaces than RRP (up to 8), will be more reliable in the event
of network outages and allows dynamic reconfiguration of interfaces.
It also fixes the ifup/ifdown and 127.0.0.1 binding problems that have
plagued corosync/openais from day 1
Signed-off-by: Christine Caulfield <ccaulfie@redhat.com>
As we now have update_node_expected_votes(), we can use that
when receiving a new EXPECTED_VOTES value from another node
rather than having our own loop.
Signed-off-by: Christine Caulfield <ccaulfie@redhat.com>
If expected_votes was set via the library but the calculation
decides it's too high, then an error is correctly returned but
the value is still set in the nodes' expected_votes field and
turns up in the corosync-quorumtool display.
This patch separates out the quorum calculation from the updating
of expected_votes per node to prevent this from happening.
Signed-off-by: Christine Caulfield <ccaulfie@redhat.com>
Reviewed-by: Jan Friesse <jfriesse@redhat.com>
