This is problematic because it can stomp over pre-existing host-specific
local files, which we'd end up removing. It also requires to run ngcpcfg
before being able to execute the tests.
We wwitch to use golang's text/template support for environment variables
so that we can stop using this hack.
Change-Id: I6c4d5b2efde2704d68a327d4df9af51bcb8fd964
- The /login/admin/ endpoint on web_int is on port 1443.
- The /login/subscriber/ endpoint on web_ext is on port 443.
- The /api/ endpoint check with user:pass on default ports should be
done on web_ext, as a similar test is already being done on web_int.
Change-Id: Ieb7f794288e52f3ac8a520d1cb3508f442648c36
These services are only supposed to be present on mgmt or li_dist nodes,
so there is no point in checking other nodes.
Change-Id: Ic7a83a30acc5078b6b2108d4393e73b7a4ce9612
Except for the mgmt nodes, each node exposes information only about
itself and its peer, so the keys and their indices depend on the nodes
involved.
Change-Id: I8562e87c246f98743d2e29e56da653da0b1b2bf6
These only make sense when they are at the end-of-line, so that no
newline gets included in the output, but never when surrounded by
other text.
Change-Id: Idd43b8e3d7de390cc1d8dc1e2c0b666012c66eba
These need to match, as the service is always going to be enabled or
disabled, depending on the conditions, in constrast to HA-handled
services, which are never enabled.
Change-Id: I1f733e3b3ccb149482e7f8577d874c3f22345724
We had to downgrade prosody in mr7.5 and have to check it is
still old/pinned version in use until we are ready for the upgrade.
See workfront ticket 66583 for more details here.
Change-Id: I7cad16c1d08beb57d52eca5f83fbcc1d21dba7c9
On early upgrades mr5.5->mr6.5 GTID was not reconfigured on
the very first upgraded node... It is a test to ensure we
have covered all the upgrade scenarios.
Change-Id: I40433f536402c52b582911fc18bd4021ed050bbf
We should not assume monit is always running and enabled, that depends
on whether it is configured as the current process_handler. This will
make it easier to remove in the future.
Change-Id: I8851728228dd3e46323d9745ed04da49b18ff928
The missing test were detected while working on another ticket.
Good to have them here as they are very important for NGCP.
Change-Id: I73afe7332dc1f60b412b9415fef1ce402829dd2b
While syncing with the Debian packaging, the binary names have changed.
Update the expected names.
Change-Id: I8c836f09898126f3749adfe15ff4b3bee59469c8
While using ngcp-licensed is not incorrect, it is an alias generated by
systemd only when the main service file is enabled. To make the check
more robust we should check for the actual canonical service name which
is ngcp-license-client.service.
Change-Id: I512563d247104474482c485ec1780bdaf32034a2
Our current policy is: all services must have .service files
(better Type=notify) to be fully integrated into ngcpcfg preset.
We need to be sure newly added services will not use systemd 'generator.late'
for automatic .service files generation.
Change-Id: Id199020be6c30f5f954496fe02c6b2fe125365fb
Otherwise we have an error here:
> root@spce:~# /usr/share/ngcp-tests/testrunner
> ...
> Error: yaml: line 3: found unknown escape character
Change-Id: I377a5796754a9310148c8361cb397e54722c8cfa
While we have lua5.1 on Debian/stretch, it's lua5.2 on Debian/buster
and lua5.3 is available as well in buster, so instead of hardcoding
the lua version 5.1, let's support whatever lua5.X version is used.
Fixes:
| not ok 34 - Command: pgrep -c -f '[l]ua5.1 /usr/bin/prosody': stdout: patterns not found: [1]
While at it, drop the unnecessary '[l]' approach, which
is a nice trick to avoid matching the grep process itself
within a `ps aux | grep ...`, but doesn't apply to
pgrep, thanks to Manuel Montecelo for spotting this leftover.
Change-Id: I3c62ddefbaf5a22f623b779426621ebaafec26b2
In CE this dir was created and used only to keep sipwise.cnf file.
Now sipwise.cnf is generated from constants.yml so no need to place it
to shared-files so there is no such dir on CE.
Add explicit check of this dir to PRO checks.
Change-Id: I35503275c8fb22b731a37747c649e9691a6674c2
Change permission of generated configs - all of them do not have 'write'
permission.
Add new sipwise_extra.cnf file to check.
Remove old non-existing
/etc/ngcp-config/shared-files/mysql/sipwise.cnf.
Change-Id: Ia08438e43783067a339de5b4cf764ce79a5c53c2
After changes to restore /etc/mysql/my.cnf to keep using "alternatives" and the
configuration moving to mariadb.cnf, as the packages coming from Debian, modify
these tests to check the new expected scenario.
Change-Id: If43c842e633151e895754f6a345472cdb9c9888c
This file is needed for when mariadb-common invokes update-alternatives with
this file as argument, otherwise it aborts the upgrade and the system is left in
a broken state.
Change-Id: Icd42ab9d0def6bc3ccbdea307f5bc6416ec96fad
The .postbuild config has been removed from /etc/mysql/grants.yml
as a result the file is installed with 0400 perms.
Fix the test accordingly.
P.S. Makefile contains 0600, but ngcpcfg removes write for everyone on apply.
Change-Id: I16830fc14c262261dc5a4c20c47bda7f296d95bc
We have not been using /var/cores for a while, since starting to use
systemd-coredump around releases mr6.2 or mr6.3, so this file is now
unnecessary.
However, there was a script creating the file and this test ensuring that it
existed and no core dump files were present.
It's not particularly tied to TT#46902, but it got discovered while working on
it and the task also touches the area of "core dump" files.
Change-Id: If75fcfa28a75359033ab4efddec503da401e61bb
The kernel.core_pattern we get now is the same being used in Debian 10
(Buster), remove the conditional.
Change-Id: I47815ef465e59d7ab058289e6d3a9ee6220d8799
kernel.sysr=1 was set for Vagrant VMs only and missing
on the real hardware AND for old releases, also it is
going to be removed from installer as well, since it
was committed to debug Proxmox issues only (review pending).
No need to test it here. Also it will fix system-tests
for upgrade 2.8->...->mr6.5 which is currently failing:
> KernelParam: kernel.sysrq: value: doesn't match, expect: ["1"] found: ["438"]
(it happens as kernel.sysrq was not set for old releases).
Change-Id: Id3b449ab9a8b48df4404233c418263029e289359
As of systemd v238 the kernel.core_pattern for
sytemd-coredump includes the hostname as well,
see systemd.git commit:
| commit f45b8015513d38ee5f7cc361db9c5b88c9aae704
| Author: Jakub Filak <jakub@thefilaks.net>
| Date: Thu Feb 15 12:12:46 2018 +0100
|
| coredump: accept hostname on command line (#8033)
Change-Id: Iac7a1723cac72848518f6eafe72c2cc2687d36a1
On newly installed CE/PRO-mr6.5/trunk the port is listened by InfluxDb.
Restored it into system tests and sent ping to Guillem for rechecking
(he is not available right now, while the test is failing and
prevent building new vagrant images).
Change-Id: I432bc9540525cf11ed343064287840a62f0f44d3
Recently subscriber self-care has been switched to new CSC by
default and the content is generated now by js framework.
https://gerrit.mgm.sipwise.com/#/c/22031/
Testing title (html title is empty now) with simple curl
doesn't work anymore. So fixing here title check to allow
goss check pass successfully.
Change-Id: I7e1c4de6d3a4027c1757590085ad76adc87251b9
This simplifies the pseudo-function calls into proper ones.
Reaname the confusing is_li_enabled into the more descriptive name
is_li_dist.
Change-Id: I42b079945c100aafa37eab9f44307fac940f42a8
We cannot keep services in background (like detached 'screen')
as systemd will kill it on client disconnect. Going to use libpam-systemd.
This reverts commit d0036830a4.
Change-Id: I52a623bd90cfa0ea8541ba9d8b9b0d10929f3ee8
Now that ssh is socket activated, and systemd is taking care of the port
listening, we need to update the address to the IPv6 ANY address.
Change-Id: I452082966e57b0981567a4150222422f7e421584
We have migrated to mariadb.service in mr6.2/systemd
while lest check both for some time for the backward compatibility.
Change-Id: Icea3acd0614ca0fd389c21a9e72687125fa2fefe
The old code didn't check the services if they were disabled in YML,
so far the service could be up and running without notice. Which is wrong.
We should test such situation.
Also move ngcpcfg-api to PRO only section as it should be not available on CE.
Change-Id: Ie4dc71e238fb6c73e26310b4ea215cf4fd3cfb1f
Let's clearly define what we are checking here, to prevent
the human mistakes due to possible .service, .path or .socket
Change-Id: I95bb7ae320fcca5655c800ac8c8281da13ee0a07
Grafana will consider the database to be R/O if it's an empty file, and
refuse to start. Check for this condition.
Change-Id: Iad2f1ed4fb03370432d438c4dcd3bab342ac2588
Empty files
/etc/mysql/conf.d/my2.cnf
/etc/mysql/inc/my.mysqld1.carrier_db.cnf
/etc/mysql/inc/my.mysqld1.carrier_non_db.cnf
/etc/mysql/inc/my.client2.cnf
/etc/mysql/client.local.cnf
/etc/mysql/client.central.cnf
are created on PRO system
Change-Id: Ide21ea301ea4ef25626c7a4ce59ea606984f43da
The output of 'asterisk -rx "odbc show"' is not a 'current' state while the
cached one, also it reports unreliable data depending on which service is
starting first asterisk or db. See the ticket for the samples here.
Change-Id: I8bbdaba077c92369d6c61711a3d6382474cd2dcb
We cannot purge some packages during migration sysv->systemd on
upgrade mr6.1->mr6.2, otherwise the final command 'reboot' will
not work as expected. It commit here can be reverted in mr7.0+.
It cannot be reverted earlier, otherwise LTS upgrade will report
sysv packages availability.
Change-Id: Ib23a046a4fae50834eddc9568f2051b556f0dcaf
In short, this script/process is unneeded in systemd-based systems, so
it has been removed.
From https://dev.mysql.com/doc/mysql-linuxunix-excerpt/5.7/en/using-systemd.html :
On platforms for which systemd support for MySQL is installed,
scripts such as mysqld_safe and the System V initialization script
are unnecessary and are not installed. For example, mysqld_safe can
handle server restarts, but systemd provides the same capability,
and does so in a manner consistent with management of other services
rather than by using an application-specific program.
Because systemd has the capability of managing multiple MySQL
instances on platforms for which systemd support for MySQL is
installed, mysqld_multi and mysqld_multi.server are unnecessary and
are not installed.
Change-Id: I1536844a3ed67e153bd986d940bfeea088e978e7
In systemd, getty is not spawned unconditionally, only when the
virtual consoled is enabled, so this process is not always running.
The config of systemd can be changed to start automatically getty
processes for the consoles and leave them waiting, but IMO changing
this just to satisfy the test is not worth it.
If in the past there have been problems with not being able to log
into the system because "getty" is not enabled maybe we can reconsider
this, or perhaps find another check that really checks if the full
login is possible (although it seems complicated), because just having
getty running is not enough.
Change-Id: Ibcfee42362457129f8d96c7c14e485de26e45083
The glusterfs has ability to listen the next port in range if
the 'base-port' is occupied by some another process.
Which is happens very often on nightly tests in Debian stretch.
I have requested goss improvement to check the port from the range:
> https://github.com/aelsabbahy/goss/issues/311
So far we have to disable the check for the moment as it is unstable:
>> goss validate
> Port: tcp:24009: listening: doesn't match, expect: [true] found: [false]
While glusterfs is OK (but listening port 24010):
> root@sp1:~# netstat -anp | grep glusterfs
> tcp 0 0 0.0.0.0:24010 0.0.0.0:* LISTEN 5412/glusterfsd
>
> root@sp1:~# gluster volume status ngcp
> Status of volume: ngcp
> Gluster process TCP Port RDMA Port Online Pid
> ------------------------------------------------------------------------------
> Brick sp1:/var/lib/glusterfs/export 24010 0 Y 5412
> Brick sp2:/var/lib/glusterfs/export 24009 0 Y 5406
Change-Id: Idaf72caf4388be9b2bf5b0e98f1d4539d9438433
Checks:
- static configuration: passing the option "net.ifnames=0" to the
kernel during boot
- check if there are such devices ("eth*") in the system when running
Change-Id: I7d17777463d4dee7c067d1bf3e08080f91898b45
- Rename the file so that it makes more sense with the new checks.
- Improve the directory contents check to be more robust against
changes in ls output (ideally goss would support this directly
with the contains directive).
Change-Id: I8bebb929240e3147627a1ec5db246f73f948799e
Guillem Jover
Mykola Malkov
Richard Fuchs
Alexander Lutay
Sergii Kipot
Michael Prokop
Manuel Montecelo