Since the user is not enabled, it cannot access the box over ssh. This
makes the usermanagement test fail, since we test the user by trying to
ssh into the box.
Signed-off-by: Matthias FRANCK <matthias.franck@softathome.com>
To be compliant with TR181 spec, the interface parameter should only take device parameters
Signed-off-by: Matthias FRANCK <matthias.franck@softathome.com>
New entries were added in newer version of the component
Output changed when switching to OpenWRT 23
Signed-off-by: Matthias FRANCK <matthias.franck@softathome.com>
* The DM of DNS.Diagnostics.NSLookupDiagnostics. is added to the tr181-dns component, test should be updated as well
Signed-off-by: Matthias FRANCK <matthias.franck@softathome.com>
Issue: Ping to wan no longer works
Cause: Firewall config was changed to no longer allow ping to wan by default
Fix: Remove wan ping test
Signed-off-by: Matthias FRANCK <matthias.franck@softathome.com>
* The default changed to now have a time server instance for every bridge defined in the network layout
* Adds the missing output for br-lcm
Signed-off-by: Matthias FRANCK <matthias.franck@softathome.com>
* Add a new test that will test toggling the instance instead of removing/adding it
* Replace the aliassed instance with a search expression to make it more robust
Signed-off-by: Matthias FRANCK <matthias.franck@softathome.com>
The default behavior is changed. We don't want to respond to ping6 WAN requests anymore.
This reverts commit 0dd60b27ebf0a61e330dc574eaa435b6a6461dec.
Signed-off-by: Matthias FRANCK <matthias.franck@softathome.com>
Instead of reducing unbound negative cache ttl try to flush it just before the tests start.
Signed-off-by: Yüce Kürüm <yuce.kurum@mind.be>
Signed-off-by: Yüce Kürüm <yuce.kurum_ext@softathome.com>
'uci_dyndns.org' server index has changed in OpenWRT 23.05. Adapt the initialization script accordingly.
Signed-off-by: Yüce Kürüm <yuce.kurum@mind.be>
Signed-off-by: Yüce Kürüm <yuce.kurum_ext@softathome.com>
Modifications to some opensource packages were made. These packages are available under a different name (suffixed with -prpl).
Signed-off-by: Matthias FRANCK <matthias.franck@softathome.com>
Netifd is not longer responsible for configuring the gateway. The TR181 components took over this task.
Signed-off-by: Matthias FRANCK <matthias.franck@softathome.com>
Since we're running DUT specific tests after generic tests, its
possible, that some of them trigger crashes in the components, so we
really need to make sure, that check-for-crashes.t test is being run as
last one, thus move it under generic/post folder and make it run after
all tests via new `CRAM_TEST_SUITE_POST` environment variable.
Fixes: 7a5c6289e6be ("ci: cram: check-for-crashes.t: make it execute as the last test")
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
There is now dedicated LCM data partition mounted under /lcm which is
not cleared up after each test, thus containers used in previous tests
persist during the DUT reboot and flashing procedure, violating test
imdepotence and making consecutive test fail.
So lets fix it by clearing up the LCM data partition during DUT firmware
flashing procedure, clearing just ext4 filesystem metadata, not erasing
complete 3 GiB and saving some time.
20 0x000f6800 0x006f67ff "lcm_data"
attrs: 0x0000000000000000
type: ebd0a0a2-b9e5-4433-87c0-68b6b72699c7
guid: c5cde408-0b88-48ba-8701-eed060c51309
The 0xf6800 is the starting eMMC block of that partition and 0x400 is
eMMC erase group size, thus minimal size we can erase.
References: PCF-1101, PCF-1267
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
As the tests are run in alphabetical order, the test is currently run as
first, but its possible, that some of the tests run in later stages
could trigger a crash and thus lets rename the filename to
zcheck-for-crashes.t and make it run as the last test.
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
Commit 45e19da4b5da ("profiles: prpl: add tr181-captiveportal and
tr181-syslog") introduced configuration changes around logging files and
their rotation, thus the standard `logread` feature stopped working as
it uses only /var/log/messages file.
So lets workaround that by using getDebugInformation command in the
meantime.
References: PCF-1316
Fixes: 45e19da4b5da ("profiles: prpl: add tr181-captiveportal and tr181-syslog")
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
Upstep of pwhm 6.13.1 brings new WiFi related data model objects
WiFi.DaemonMgt and WiFi.DaemonMgt.Daemon for single instance hostapd
support.
Update of UPDK to version 9.1.65 for OSPv2 WiFi7 capable board does
support single instance hostapd already, thus WiFi.Vendor.Daemon object
is added in OSPv2 vendor module data model, thus we can't use generic
ubus datamodel list test (generic/ubus-datamodel-list.t) anymore on this
board, so the new OSPv2 board specific test was created
(mxl25641-hdk-6/ubus-datamodel-list.t).
References: PPMR-226, PPM-2814
Signed-off-by: Bhavesh Bharadiya <bnbharadiya@maxlinear.com>
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org> [CI tests fixes, commit description]
PCF:1229 issue is fixed in UPDK 9.1.65 WLAN version 06.01.00.2047.
Now when VAPs are down, no hostapd instance is running.
Correcting ci tests for OSPv2 platform, removing the workaround.
Signed-off-by: Bhavesh Bharadiya <bnbharadiya@maxlinear.com>
Update mxl_wlan_hostap_ng_wav700 profile for below changes:
* bump to version 06.01.00.2047
Update mxl_x86 OSPv2 profile for below changes:
* bump to version updk_9.1.65
* Flash layout for OSPv2 is prpl Standard Flash layout v1.0 compliant
Fixes: PCF-1229
Signed-off-by: Bhavesh Bharadiya <bnbharadiya@maxlinear.com>
Bump feed-prpl to bring in following changes:
* prplMesh: pwhm: bump to v6.13.1
* pwhm version 6.13.1 fixes several issues
UPDK 9.1.65 mod_whm (MxL vendor module) has dependency on pwhm v6.13.1
APIs hence it is needed to bump version.
Signed-off-by: Bhavesh Bharadiya <bnbharadiya@maxlinear.com>
obuspa, as integrated into prplOS, contains a vendor patch that waits
for the preconditions `Device.DeviceInfo` and
`Device.Time.Status==Synchronized` before attempting to connect to the
USP broker. These preconditions are currently mandatory and cannot be
disabled.
This update fetches the latest version of obuspa from the BBF v8.0.4
prpl-integration3 branch, which adds support for a timeout, controlled
by the parameter `Device.LocalAgent.X_VANTIVA-COM_PreConnectTimeout`.
The default timeout is 60 seconds. This timeout functions as a fallback
mechanism, allowing obuspa to try to connect to the USP broker even if
the preconditions are not met.
References: PCF-1104
Signed-off-by: Sandesh Mahale <sandesh.mahale@vantiva.com>
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
(cherry picked from commit 64cc6d328171db2a46485e1e2a8a188e055d5c6e)
Upstream lighttpd defaults to mbedTLS for its internal cryptolib
handling as its default TLS provider, but for prplOS it was decided to
stick to only single TLS/crypto provider which is OpenSSL, so lets make
it default for lighttpd as well.
References: PCF-1006, PCF-1182
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
This reverts commit 261489bfa60e8a54f691240e9a2fe8161baf2234 as the
issue was fixed upstream.
References: PCF-1182, PCF-1006
References: https://github.com/openwrt/packages/pull/24423
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
In commit 5bbd04eaa47a ("PROVISORY: ci: cdrouter: attempt to workaround
NTP sync issues") we've increased CDRouter test startTimeout from 90 to
180 seconds, but it turned out to be not enough, so lets increase
CDRouter startTimeout even further to 360 seconds and see how it
goes.
References: PCF-1168, PCF-1257
Suggested-by: Vincent Harle <vincent.harle@softathome.com>
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
(cherry picked from commit abcb4d212652b3c0dc4bb2db05640a7705bd6f47)
In the commit 3399c5f970c0 ("profiles: prpl: add
CONFIG_BUSYBOX_CONFIG_FLOAT_DURATION config option") we've introduced
init scripts which rely on Busybox's sleep command having support for
float duration, but we've forget to assured that on QA, so lets add the
test now.
References: PCF-1151, PCF-1260
Fixes: 3399c5f970c0 ("profiles: prpl: add CONFIG_BUSYBOX_CONFIG_FLOAT_DURATION config option")
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
(cherry picked from commit f5a5f9d536c81c07278d44621ee6f9bde87f27a0)
time-manager's managed chrony should be the sole system component
responsible for time management. Currently, Busybox's `ntpd` is also
running, attempting to perform the same task, which may cause
undesirable system behavior.
This issue became more apparent after commit 568acaccf8b4 ("ci:
cdrouter: fix unable-to-start-connection-to-usp-agent by enabling NTP
sync"), where both Chrony and ntpd began using the same NTP pools,
0.europe.pool.ntp.org and 1.europe.pool.ntp.org.
To address this, we should make `time-manager` exclusively responsible
for system time management, remove Busybox's `ntpd`, add a CI check to
prevent this regression in the future.
References: PCF-1257, PCF-1104
Suggested-by: Vincent Harle <vincent.harle@softathome.com>
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
(cherry picked from commit 8e1f17f34b3c9a0b174a7b22df0be889390a637b)
Currently the Lua script `boot_measurements.lua` collects quite handy boot
time metrics, but unfortunatelly those metrics are not available via
getDebugInformation script, thus missing in most bug reports. So lets
fix it by hooking in the standard init script `debuginfo` command which
would dump the `/tmp/amx_mon_objects.log` into the output.
References: PCF-1168, PCF-1242, PCF-1257
Upstream-Status: Submitted [https://gitlab.com/prpl-foundation/components/utilities/debug-information/-/merge_requests/3]
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
(cherry picked from commit 4df1744a5f73bca474af6d1d7b9701d3ee8d337f)
With CONFIG_BUSYBOX_CONFIG_FLOAT_DURATION option we can sleep for
a fractional duration. This increases the granularity of control where
sleep used for periodic tasks.
References: PCF-1151
Signed-off-by: Yüce Kürüm <yuce.kurum@mind.be>
Signed-off-by: Yüce Kürüm <yuce.kurum_ext@softathome.com>
(cherry picked from commit 3399c5f970c05b63b4e53be52d22b508429fed28)
With the introduction of new shutdown function, init scripts started
wait until the component shutdown process is finished. This has the
unwanted side effect of stopping the components one by one during reboot
and shutdown.
Move component shutdown processes to background and wait with
amx-shutdown-wait script until all of them finishes or a timeout occurs.
Fixes: PCF-1151
Signed-off-by: Yüce Kürüm <yuce.kurum@mind.be>
Signed-off-by: Yüce Kürüm <yuce.kurum_ext@softathome.com>
(cherry picked from commit 60a37a7c6414254d8c64d18b7bc14ad63b7be02b)
* obuspa: decrease the log verbosity to 2=Warning
Currently the default log settings is set to 3=Info which is a bit
chatty and usually only needed during debugging sessions. Thus lets
decrease the logging verbosity to 2=Warning which should be good enough
during normal system operation. It can be enabled quite easily when
needed via `obuspa -c verbose 3` shell command.
* obuspa: disable default very verbose USP protocol tracing
Currently there is USP protocol tracing enabled which leads to quite
huge amount of logs which are usually only needed during debugging
sessions. Thus lets disable this logging as it can be enabled quite
easily when needed via `obuspa -c prototrace 1` shell command.
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
(cherry picked from commit 60dcb7a9c86084c8711bd33d48116b000630644a)
It was found out, that more time is needed to make HL-API certification
over USP passing, thus we're disabling the automatic testing and turning
it into manual, so we don't waste resources on failing test cases, but
keeping the test so we can run it manually if needed be.
References: PCF-1095, PCF-1218, PCF-1252
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
(cherry picked from commit bf72af976bcd01ea7e0cb50be110f9a57e88fcc3)
Currently obuspa get Device queries take an increasing time with every
query. This causes CDRouter USP tests to timeout.
Let's increase the USP message timeout until a proper fix is introduced.
References: PCF-1252
Suggested-by: Vincent Harle <vincent.harle@softathome.com>
Signed-off-by: Yüce Kürüm <yuce.kurum@mind.be>
Signed-off-by: Yüce Kürüm <yuce.kurum_ext@softathome.com>
(cherry picked from commit 4f3dd6db11964bde5ba34cb9b9b4f7c1af71a603)
Currently we're experiencing some NTP time sync issues:
54.82 - IP interface wan has ipv4address = 10.0.0.2 (IP.Interface.2.Status = Up, IP.Interface.2.IPv4Address.1.Status = Enabled)
...snip...
59.64 - Router main has gateway ipaddress = 10.0.0.1
...snip...
113.35 - Time Client cpe-client-1 is Synchronized
114.45 - Time Client cpe-client-1 is Unsynchronized
...snip...
174.46 - Time Client cpe-client-1 is Synchronized
NTP sync was achieved at May 7 11:50:50
May 7 11:50:50 prplOS obuspa[2025]: resolved_path_results {
May 7 11:50:50 prplOS obuspa[2025]: resolved_path: "Device.Time."
May 7 11:50:50 prplOS obuspa[2025]: result_params {
May 7 11:50:50 prplOS obuspa[2025]: key: "Status"
May 7 11:50:50 prplOS obuspa[2025]: value: "Synchronized"
May 7 11:50:50 prplOS obuspa[2025]: }
May 7 11:50:50 prplOS obuspa[2025]: }
and within 3 sec testing is stopped
May 7 11:50:53 prplOS CI: Testing finished, up 5 minutes
Looks like chronyd was able to get time sync from 3 sources:
May 7 09:00:54 prplOS chronyd[2953]: Selected source 3.3.3.7 (1.europe.pool.ntp.org)
May 7 11:49:45 prplOS chronyd[2953]: Selected source 3.3.3.6 (0.europe.pool.ntp.org)
May 7 11:49:50 prplOS chronyd[3916]: Selected source 3.3.3.7 (1.europe.pool.ntp.org)
OBUSPA polls Device.Time.Status every 10sec, query at 11:50:40 returns status as Unsynchrozied
May 7 11:50:40 prplOS obuspa[2025]: resolved_path_results {
May 7 11:50:40 prplOS obuspa[2025]: resolved_path: "Device.Time."
May 7 11:50:40 prplOS obuspa[2025]: result_params {
May 7 11:50:40 prplOS obuspa[2025]: key: "Status"
May 7 11:50:40 prplOS obuspa[2025]: value: "Unsynchronized"
May 7 11:50:40 prplOS obuspa[2025]: }
Thus status was not updated atleast till 11:50:40.
So lets increase current CDRouter startTimeout from 90 to 180 seconds
and see how it goes.
References: PCF-1168
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
(cherry picked from commit 5bbd04eaa47aed4b077e5403ea7173e310e1f779)
Current CDRouter tests are not passing due to the following issue:
FAIL: The USP agent failed to acknowledge the initial USP message, resulting in the error: "unable-to-start-connection-to-usp-agent".
The failure is occurring because the device is unable to obtain NTP
time. The status of Device.Time is listed as Error_FailedToSynchronize.
Under task PCF-1104, OB-USP-A is configured to withhold attempting
connections to the ACS until essential USP services like DeviceInfo. are
registered, the WAN connection is established, and NTP time is
successfully synchronized. In this scenario, the NTP time has not been
synchronized, which is causing the error.
So lets unblock this by enabling the NTP time synchronization by using
the default/pre-configured prplOS 0.europe.pool.ntp.org and
1.europe.pool.ntp.org NTP servers.
References: PCF-1104
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
(cherry picked from commit 568acaccf8b4fabcdaf66190d74048425d7d7043)
