Age | Commit message (Collapse) | Author | Files | Lines |
|
Currently we trigger the phosphor-debug-collector application to create
a dump whenever a file is present in the pstore directory. This means we
create a dump if eg. a console or ftrace log is captured by ramoops,
which is not indicative of an error.
For example, this is the content of the ramoops filesystem after a
normal reboot followed by a crash:
console-ramoops-0 dmesg-ramoops-0 dmesg-ramoops-1
Change to ConditionPathExistsGlob and look for anything with the pattern
dmesg-ramoops-*.
This is a better fix for the workaround implemented in 17e10e646324
("meta-ibm:pstore: remove console from ramoops"), allowing systems to
collect more traces without invoking a dump.
Change-Id: If1d93ee86a2cfb7f0902989cb78d128e82578402
Signed-off-by: Joel Stanley <joel@jms.id.au>
|
|
- Create a new package for the ramoops service and only load it once
when the debug-collector daemon starts
- Fix openbmc/phosphor-debug-collector#15
Tested:
- Built obmc-phosphor-image and saw the ramoops-monitor run
successfully.
Signed-off-by: George Liu <liuxiwei@inspur.com>
Change-Id: Ic616408b834a257a981f0e0ef24ad030ccffc616
|
|
phosphot-dump-monitor adds watch on systemd core dump path to
detect any new core files created to initiate a dump. On some
systems systemd core dump package is not pulled, due to this
phosphor-dump monitor service logs error and tries retry for 3 times.
Modified to start phosphor-dump-monitor service only when the
systed core dump path exists.
Tried to add Requires on systedm core dump service but as it is
a template version going with path approach.
Tested:
1) Removed systemd dump folder and restarted dump-monitor service
and ensured that there are no errors logged
(From meta-phosphor rev: 7a99220eaae763113372477c3d25039284802ef7)
Signed-off-by: Marri Devender Rao <devenrao@in.ibm.com>
Change-Id: Ibeb83cde454bf4c6fc6c6827a25705649ceaabe5
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
A number of corequisites exist so squashing these all into one patch.
meta-ingrasys - refresh master bc513127f4..67bcaf389f
meta-hxt - refresh master 86a4df514f..8fe0d38bbb
meta-phosphor - refresh master 06c09d7cce..df6ddae3d2
meta-quanta - refresh master 4f126361a1..53f3025271
Change-Id: I3daeef2a0467d8ea4f1fd1c617e526f8b11258af
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
It is observed that systemd-coredump has the capability to remove
the older dumps which are having the naming style mentioned in
/proc/sys/kernel/core_pattern, But will not remove any other files
present in the coredump directory.
It can happen that, while processing a dump the BMC Crashes and there
by leaving the partial dumps in the coredump directory and these will
be hidden and will not be removed by systemd-coredump service.These
files can get piled up and the BMC can run into OOM condition.So idea
of this commit is to make sure that , we remove all the temporary
hidden files in the coredump directory during the next BMC boot.
- Made a rule and placed it in /usr/lib/tmpfiles.d/*.conf which will
be used by systemd-tmpfiles to remove the hidden files in coredump
directory during boot.
Partially Resolves openbmc/openbmc#2826
(From meta-phosphor rev: 9fe03c666476de51ab75904f008f45f8485ac9b2)
Change-Id: Iff96b3880ebdeccb842e6aa7ff9807fcd451fb5a
Signed-off-by: manojeda <manojeda@in.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Adopt a more conventional directory hierarchy. meta-phosphor is still
a _long_ way from suitable for hosting on yoctoproject.org but things
like this don't help.
(From meta-phosphor rev: 471cfcefa74b8c7ceb704cb670e6d915cf27c63b)
Change-Id: I3f106b2f6cdc6cec734be28a6090800546f362eb
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|