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.
Defaulting to host public IP addresses when all nodes are unhealthy
does not obey the principle of least surprise. It has caused much
confusion over the years. It often leads to problems when all nodes
are unhealthy due to something like a cluster filesystem being
unmounted.
Change the default value for this tunable as the first step of
completely removing this behaviour.
Remove tests that set NoIPHostOnAllDisabled=1 and update the expected
result for other tests where no nodes are healthy.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Although this tests correctness it is most useful for testing that
changes to IP takeover algorithm do not cause obvious performance
regressions.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Autobuild-User(master): Amitay Isaacs <amitay@samba.org>
Autobuild-Date(master): Thu Aug 10 10:30:58 CEST 2017 on sn-devel-144
At the moment this is done as a post-processing step for any nodes
that have no known IPs. However, this doesn't allow testing of
scenarios where there no known IPs on one or more nodes.
Add relevant tests.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
It isn't used anywhere and doesn't contain some of the optimisations
that have since gone into the C code.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Autobuild-User(master): Amitay Isaacs <amitay@samba.org>
Autobuild-Date(master): Mon Dec 19 07:58:45 CET 2016 on sn-devel-144
Instead of gathering the value from all nodes, just use the value on
the recovery master and have it affect all nodes.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Instead of gathering the value from all nodes, just use the value on
the recovery master and have it affect all nodes.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
This is now achieved by defining functions extra_header() and
extra_footer().
Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Use a variable to allow easy change of this string in case future
logging changes modify the timestamp format or do not support
timestamping.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Volker Lendecke <vl@samba.org>
srcimbl gets changed on every iteration of the loop. The value that
should be stored for the new imbalance of the source node is
minsrcimbl.
To help diagnose this, added some extra debug that can be left in.
The extra debug changes the output of a couple of tests. Note that
the resulting IP allocations in those tests is unchanged - only the
debug output is changed.
Also add some new tests that illustrates the bug.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Using a variable is too fragile, so use a function instead.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Michael Adam <obnox@samba.org>
2 tests to show a bad result and a 3rd test for the fix.
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit ef35c8889d90220929e48e66eb62da9ea2025ede)
Signed-off-by: Martin Schwenke <martin@meltin.net>
Pair-programmed-with: Amitay Isaacs <amitay@gmail.com>
(This used to be ctdb commit 7cf63722873a6a7baafd77aa3d8a1989b221dee9)
This really needs to be per-node. The rename is because nodes with
this tunable switched on should drop IPs if they become unhealthy (or
disabled in some other way).
* Add new flag NODE_FLAGS_NOIPHOST, only used in recovery daemon.
* Enhance set_ipflags_internal() and set_ipflags() to setup
NODE_FLAGS_NOIPHOST depending on setting of NoIPHostOnAllDisabled
and/or whether nodes are disabled/inactive.
* Replace can_node_servce_ip() with functions can_node_host_ip() and
can_node_takeover_ip(). These functions are the only ones that need
to look at NODE_FLAGS_NOIPTAKEOVER and NODE_FLAGS_NOIPHOST. They
can make the decision without looking at any other flags due to
previous setup.
* Remove explicit flag checking in IP allocation functions (including
unassign_unsuitable_ips()) and just call can_node_host_ip() and
can_node_takeover_ip() as appropriate.
* Update test code to handle CTDB_SET_NoIPHostOnAllDisabled.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Pair-programmed-with: Amitay Isaacs <amitay@gmail.com>
(This used to be ctdb commit 1308a51f73f2e29ba4dbebb6111d9309a89732cc)
Implemented for CTDB_SET_NoIPTakeover.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Pair-programmed-with: Amitay Isaacs <amitay@gmail.com>
(This used to be ctdb commit a1addd89fd9c0390912604097acd028cc24d3483)
Neither basic_failback() nor lcp2_failback() unassign IPs anymore, so
there's no point looping back that far.
Also fix a unit test that now fails because looping back to handle
unassigned IPs is no longer logged.
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit c09aeaecad7d3232b1c07bab826b96818756f5e0)
3 tests should assign IPs to all nodes.
3 tests set NoIPTakeoverOnDisabled=1 and should drop all IPs.
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit edda58a45915494027785608126b5da7c98fee85)
Default to LCP2, like ctdbd. Also support "det" for deterministic
IPs.
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit 20631f5f29859920844dd8f410e24917aabd3dfd)
Move some options from eventscripts/run_tests.sh to scripts/run_tests.
Remove the former.
Move some functions from eventscripts/scripts/local.sh to
scripts/unit.sh.
Both of these are modified during move so they are no longer
eventscript-specific.
Tweak */local.sh so that the new functions in unit.sh are used.
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit 7ff485687891732074c9fc9998502ca197663d02)
A recent change in the deamon's (i.e. real) IP allocation algorithm
uses qsort(3). This makes it very difficult to get the same results
in Python, since qsort's behaviour on identical keys is undefined.
Add a -e option to run "../bin/ctdb_takeover_tests ctdb_takeover_run_core"
instead of the internal Python algorithm. We pass the current cluster
state and read it back from the output of the test program.
This is currently a rough hack. Perhaps it'll be improved later...
:-)
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit eed3f4988290bbe317a7aaa6f61826676dd8365a)
The default in this script is still deterministic IPs. The LCP2
option should really turn off deterministic IPs.
Also fix a typo...
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit c378a059b35fcc9955215a0095826c919d42153c)
A recent fix made the LCP2 algorithm try harder find a candidate
source node. The debug output shows extra output because it is trying
harder so we accommodate that.
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit 9a6f918bf6db79c1f8c53c0df23d47b73c117ea2)
There are 4 IPs across 2 nodes and 2 addresses across 2 other nodes.
If one of the latter nodes is unhealthy and then becomes healthy
again, an IP isn't failed back. This is because the nodes in the 1st
group are >= unbalanced then the nodes in the 2nd group.
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit 1d26e7cff6292fcbf63efc4628ffbb63b6f1f73c)
This makes IPv4 addresses comparable with IPv6 but reduces the overall
effectiveness of the algorithm. The alternative would be to treat
these addresses separately while trying to keep all the IPs in overall
balance... which is basically the problem that LCP2 solves. :-)
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit 3a7624f9d468b99714a7b6a45313f9e7f66011ed)
This time in the stats summary.
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit aabb2507dacc63ae026e6c99704a2fb79950e82c)
When there are IP groups, do not terminate when the overall cluster
goes out of balance.
Also make explicit that grat_ip_moves is an integer not a boolean, so
only terminate if it is greater than 0.
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit 0899f14b1483682d73d1ee2d2419db54ffeadc4b)
It had out-of-date information and a typo.
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit 5d0d2b8b528414c859da0e6fd5959321db33608b)
This is likely to cause many more state changes for nodes. In this
mode the odds of a failover are applied to determine whether a state
change occurs for each node. If no state change occurs then the
process is repeated.
Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit b7c42bff9457ec8294b04245af8e3b6010707d1a)
