| Project Key | Project Name | PTL | Description |
|---|---|---|---|
| RICAPP | RIC Applications | SUNIL SINGH | This includes open source sample xAPPs and platform applications that can we used for integration, testing, and demonstrations. |
| RIC | Initial RIC Platform to support xAPPs with limited support for O1, A1, and E2 interfaces. aa | ||
| OCU | O-RAN Central Unit | TBD | The OCU should target an initial software deliverable with limited functionality. Focus on aa basic E2 interface to enable initial integration testing between RIC and OCU. |
| ODUHIGH | Manasi Padhy | Focus on initial L2 functional blocks based on seed code contributions. | |
| ODULOW | O-RAN Distributed Unit Low Layers | Focus on initial L1 functional blocks based on seed code contributions. | |
| ORU | O-RAN Radio Unit | TBD | Out of scope for release "A". |
| OAM | Operations and Maintenance | O1-related sysdev; initial dashboard for monitoring and demonstration of contributions. More ... | |
| SIM | Simulations | Alex Stancu | Initial simulators used for testing O-RAN NF interfaces. |
| INF | Infrastructure | Initial building blocks for infrastructure to run O-RAN NF components. | |
| INT | Integration and Testing | The integration and test effort will focus on testing the requirements documented in each release. This will focus on end to end test and use case testing. NF Testing shall primarily be done on the Akraino-REC (Radio Edge Cloud) blueprint. While Non-RT RIC and O1 interface testing will primarily use ONAP components. | |
DOC | Documentation | weichen ni | Documentation on contributions and O-RAN SC contents. |
| NONRTRIC | Non-RealTime RIC (RAN Intelligent Controller) | Non-Realtime RIC focus on feature functionality of A1-interface, R1 interface, and support rApp (applications). More ... | |
| SMO | Service Management and Orchestration (SMO) | The primary goal of the SMO project is to integrate different software artifacts of existing open-source projects creating a fully functional open-source Service Management and Orchestration (SMO). | |
| AIMLFW | AI/ML Framework | hoejoo lee | Initial AI/ML workflow implementation for O-RAN environment. Need to interact with another project to accomplish a whole life cycle management of the AI model. |
NOTE: Repositories that are indicated as scp/ repositories are repositories maintained by the O-RAN Alliance and licensed under the O-RAN Specification Code Project under licenses that are different from the licensing of the O-RAN Software Community Project. For questions concerning an O-RAN Specification Code Project code or documentation, please contact the O-RAN Alliance.
54 Comments
Jun Hyuk Song
Great! I am looking forward to hearing more from O-RAN-OSC.
Farheen Cefalu
John Keeney , there are two entries in the Projects table above for Non real time RIC. One at the top and one at the bottom. Could you please consolidate them into one?
Rakesh Mhapsekar
whether reporting will also be part of RIC application project? I believe reporting shall also be considered along with testing, Integration and demonstration.
Moez Mansour
Dear Experts,
this is my first time i join an open source project community in fact i'm a RAN Eng working on 4G/5G vendor Equipment (Huawei/Ericsson) KPI, Parameters setting and troubleshoting QoS issue, i'm very interested by Open RAN project i would like to participate in the efforts of making RAN network more open, if any one can help me with the IT part (Kubernetes,json) so i ca join the adventrure with you i will be very grateful
Thanks in advance and best regards
Moez
Mahesh Jethanandani
Did you mean O-RAN or OpenRAN(TIP)? They are different open source communities.
Martin Skorupski
Hi Moez,
welcome to the party
- for the k8s part on SMO/OAM level you may want to take a look here: Deployment
Please let us know if you are looking for something else (e.g. k8s for O-DU?!?)
Happy coding,
Martin
Claudio Ferreira Dias
Hello Everyone!
I am looking for a solution to deploy a Radio Access Network. It seems that Open RAN is the way to go. I am still figuring out how all 5G pieces are connected, but it isn't clear. I think most of my problems are solved by answering the following questions:
Thanks for your time. I will appreciate it a lot if my questions are solved.
Manasi Padhy
Hello,
Thank you for the interest in O-RAN open source, I may like to answer few of your queries,
Sindhu Pitta
Hi,
I am trying the pairwise testing of ODU high and ODU low and was able to compile and build the dependencies.
But while running/executing the binaries facing issues.
Facing issue in running l1 binary in radio mode.
the error is like " couldn't find the 'phycfg_xran.xml' file", even checked for that file in this directory "FlexRAN/l1/bin/nr5g/gnb/l1/" but couldn't find it.
We could see the message flow between odu-h (testmac) and odu-l when we ran L1 with TESTMAC. We have run the testmac script "https://docs.o-ran-sc.org/projects/o-ran-sc-o-du-phy/en/latest/run_l1.html#run-l1-with-testmac "
We should see the p5/p7 message flows when we run odu-h binaries, but we couldn't see it. When we run odu-h binaries, we could see the message saying L2 is up below output is seen
[INFO_LOG]L2 is up...
wls_fapi_create_mem_array: pMemArray[0xaebe50] pMemArrayMemory[0x4400000000] totalSize[1052770304] nBlockSize[262144] numBlocks[4016]
[INFO_LOG][FAPI_INT] WLS init Successful
[INFO_LOG]ORAN_5G_FAPI init successful
ORAN_5G_FAPI>[INFO_LOG][PHY2MAC] Thread nr5g_fapi_phy2mac_thread_func launched LWP:139930319263488 on Core: 12
[INFO_LOG][MAC2PHY] Thread nr5g_fapi_mac2phy_thread_func launched LWP:139930310870784 on Core: 11
But when we run the odu-h binaries with timer mode or radio mode, we didn't see any message flow btw odu-h and odu-l.
Is there any config change required or any other option is required to see the message flow wrt pairwise testing?
https://docs.o-ran-sc.org/projects/o-ran-sc-o-du-l2/en/latest/overview.html#o-du-high-interfaces - P5 and P7 messages are mentioned here
https://docs.o-ran-sc.org/projects/o-ran-sc-o-du-l2/en/latest/user-guide.html#iii-execution - execution
Regards,
Sindhu pitta.
Perapalla Maneesha
Hi,
We are trying to run sample application of DU and RU following the link https://docs.o-ran-sc.org/projects/o-ran-sc-o-du-phy/en/latest/Setup-Configuration_fh.html#a-1-setup-configuration.
While we are launching the application we are getting Error: "Loading file ./usecase/mu0_10mhz/ant_0.bin to DL IFFT IN IQ Samples in binary format: can't open file ./usecase/mu0_10mhz/ant_0.bin!".
We learnt from oran documentation on ODU-L that Octave software is required for generating IQ samples for that.
Can someone guide us on how to generate those samples using Octave.
Thank you,
Perapalla Maneesha.
Sindhu Pitta
Hello Everyone,
I am integrating ODU Low +ODU High +CU +Free5GC for the end-to-end testing.
I was able to connect Free5GC+CU+ODU High and test the registration procedure by following the below-given steps to bring up free5gc
https://github.com/free5gc/free5gc/wiki/Installation
But when connecting the whole integrated setup with ODU Low I am unable to see the registration procedure and can see errors in Fapi and L1 binaries
Error in Fapi binary :
Error in l1 binary :
but I can see the p5 messages between ODU High and ODU low\
Logs in ODU High :
The following link is the one I am using for running the ODU-H, L1, Fapi, CU :
https://docs.o-ran-sc.org/projects/o-ran-sc-o-du-l2/en/latest/user-guide.html#b-pairwise-testing-with-intel-o-du-low
Can someone guide us in how to check the registration procedure wrt ODU Low?
Looking forward to the solution.
Thank You,
Sindhu Pitta.
Harshita Lal
Hi Sindhu
From ODU-High log snippet, I see that CONFIG.Request and START.Request are successful and the very first SLOT.Indication from L1 has reached to ODU-High. Hence the Cell is marked as Active.
Post this, "API Ordering is wrong" error is observed at FAPI Translator module. Due to which L1 binary stops and sends STOP.Indication to ODU-High.
In order to check events occurring between first SLOT.Indication and this error, please increase the log level to collect a more detailed logs at FAPI Translator module. You can do the same by changing "level = info" to "level = trace" in file "<O-DU Low Directory>/phy/fapi_5g/bin/oran_5g_fapi.cfg".
Also, could you please confirm if the latest ODU-High binary is used in the above testing.
Regards
Harshita
Sindhu Pitta
Hi Harshita,
Thanks for your reply.
We are using the ODU-H from the Cherry release.
We have changed the log level to trace and we are building odu with the options: make odu PHY=INTEL_L1 MACHINE=BIT64 MODE=FDD
Fapi logs:
We see the infinite loop running in fapi .
We tried building odu with 'make odu PHY=INTEL_L1 MACHINE=BIT64 MODE=FDD PHY_MODE=TIMER' and we are running l1 binary with the option './l1.sh -e',
Attaching the ODUand L1 logs as a file below.
Now we are not seeing the error as 'API Ordering is wrong', but we the infinite loop running in FAPI and we could see the logs:
[TRACE_LOG][FAPI2PHY WLS][GET] 0x2f8d080000 size: 48 type: 10 flags: 0
[DEBUG_LOG]To Free 1
[DEBUG_LOG]Free 1
[TRACE_LOG][PHY2MAC] nr5g_fapi_phy2mac_api_recv_handler:
[DEBUG_LOG][FAPI MSG HDR] FAPI Message Header Added for PHY: 0
[ERROR_LOG][SLOT.indication] Unable to create list element. Out of memory!!!
Currently, we don't have any UE connected. The current setup is: Free5gc+CU_stub+ODU-H+ODU-L+RU simulator
When we tried
l1-1
odu-1Free5gc+CU_Stub+ODU-H, we could see the free5gc registration is happening.
But when we connected to ODU-L and launched binaries, we don't see the free5gc registration is happening.
Harshita Lal
Hi Sindhu,
Thanks for the detailed logs.
In Cherry release, the binaries were tested only in Timer mode whereas testing in Radio mode was done in later releases. Some fixes were done at ODU-High to make it compatible with ODU-Low running in Radio mode.
Hence, we request you to test with the latest ODU-High binary i.e. E-Release binary.
Regards
Harshita.
Sindhu Pitta
Hi Harshita,
Thank you for the info.
when we run in radio mode we see the error like phycfg_xran.xml file is not found.
When we check in Flexran codebase then also we couldn't find the mentioned file.
Thank you,
Sindhu pitta.
Sindhu Pitta
Hi Harshita,
So, as mentioned above while running in radio mode we were facing a file not found error.
We modified the phycfg_timer.xml file ran in radio mode and captured the FAPI, L1, and ODU-High logs.
Please find the attached three files and can guide us on how to proceed further.
Thank you,
Sindhu Pitta.
Harshita Lal
Hi Sindhu
From above logs, L1 seems to have started processing START.Request but there was no START.Response to FAPI Translator module.
I suspect this could be because modified phycfg_timer.xml is used instead of phycfg_xran.xml which has all the radio mode configurations.
For "configuration file phycfg_xran.xml is not found" error and other ODU-Low related queries, please contact Luis Farias.
Thanks & Regards
Harshita
Sindhu Pitta
Hi Luis Farias and Harshita,
We are using,
1. Free5GC Core (https://github.com/free5gc/free5gc)
2. Our own CU SIMULATOR
3. Radysis DU(E-release) as ODU-H (github.com/o-ran-sc/o-du-12)
4. O-RAN DU Physical Layer ( https://gerrit.o-ran-sc.org/r/o-du/phy.git)
5. Open source FlexRAN (https://github.com/intel/FlexRAN)
With TIMER mode:
When we integrated 1, 2, and 3, we could achieve the free5gc registration successfully.
But When we integrated with Flexran L1 (we are using opensource flex ran), we couldn't make the free5gc registration happen. L1 seems to have started processing
'START.Request' but there was no START. Response to FAPI Translator module. Attached are the logs with the timer option.
With Radio mode:
To run L1 with radio mode, we don't find the phycfg_xran.xml file as part of the
FlexRAN opensource package. We tried using phycfg_timer.xml only for radio mode by
updating the <XRAN> config options.
But we see the same results as above(attached logs with radio option).
Could you please help us to proceed further?
Thanks& Regards,
Sindhu Pitta.
Rajisha Payyanur
Hi Team,
We have integrated ODU (Radysis's ODU-H and Opensource Flexran) with External RRU. And we tried launching sample-app on DU.
We could see,
Can anyone please suggest how to tune the power scaling parameters at DU to send the IQ data with the expected dbfs level.
Regards,
Rajisha
Sindhu Pitta
Hi Team,
We would like to get a clarification on O-CU used in E2E testing.
In one of the demo notes, I could see that Radysis's O-CU is used along with intel and Radysis's Du.
Is this CU_STUB which we get along with DU high or full-fledged O-CU?
Could anyone share the details?
Regards,
Sindhu Pitta.
Sindhu Pitta
Hello,
can someone please update me on the status of the O-CU component?
I can only find this related repository - https://github.com/o-ran-sc/scp-ocu-5gnr.
Is there a more updated codebase somewhere and is there any documentation I can follow?
Thank you,
Sindhu Pitta.
David Kinsey
Sindhu, we use the Radysis Binary which we obtained a license from Radysis to use. The OCU project code base is old, incomplete as an NE, and deprecated since the project has no PTL or contributors. The OCU is not a stub in which case it needs to be connected to a 5G Core. So your environment would also need an AMF/UPF. We use a core emulator from Viavi. I am not sure but I think the CU_STUB is a stand-alone test stub to mimic the needed OCU interface needed for ODU-HIGH functional testing.
Perapalla Maneesha
Hi Team,
We have cloned new phy code i.e oran_e_maintenence_release_v1.0. We are facing issues while building libraries.
While building wls library we are getting error:
./build.sh
Building dpdk based wls library
Package libdpdk was not found in the pkg-config search path.
Perhaps you should add the directory containing `libdpdk.pc'
to the PKG_CONFIG_PATH environment variable
No package 'libdpdk' found
============================================================================================
Building libwls.so
============================================================================================
[BUILD] lib : libwls
[CC] wls_lib_dpdk.o
wls_lib_dpdk.c(34): catastrophic error: cannot open source file "rte_eal.h"
#include <rte_eal.h>
^
compilation aborted for wls_lib_dpdk.c (code 4)
make: *** [make/obj/wls_lib_dpdk.o] Error 4
Package libdpdk was not found in the pkg-config search path.
Perhaps you should add the directory containing `libdpdk.pc'
to the PKG_CONFIG_PATH environment variable
No package 'libdpdk' found
============================================================================================
Building wls_test
============================================================================================
[BUILD] elf : wls_test
[CC] pool.o
[CC] testapp.o
testapp.c(41): catastrophic error: cannot open source file "rte_config.h"
#include <rte_config.h>
^
compilation aborted for testapp.c (code 4)
make: *** [make/obj/testapp.o] Error 4.
We have followed the steps given in O-RAN-SC Docs (https://docs.o-ran-sc.org/projects/o-ran-sc-o-du-phy/en/latest/build_prerequisite.html)
to install the prerequisites including dpdk. But still facing dpdk related issues while building wls library.
Can anyone please let us know the build prerequisites for building the new phy code libraries.
Regards,
Perapalla Maneesha.
Sindhu Pitta
Hi Harshita Lal.
we are trying to run fapi with the new release from:
https://gerrit.o-ran-sc.org/r/gitweb?p=o-du/phy.git;a=commit;h=2de97529a4c5a1922214ba0e6f0fb84cacbd0bc7
but seeing the below error:
Can you please help with this?
Thank you
Regards,
Sindhu pitta.
Harshita Lal
Hi Sindhu
From the logs you shared, an issue seemed to have occurred during the initialization itself of FAPI translator module. At this point, there is no communication between FAPI translator and ODU-High yet.
Please contact Luis Farias form Intel team for the same.
Regards
Harshita.
Luis Farias
Sindhu & Perapalla: The documentation for the oran_e_mainteance_release has not been autoloaded to the oran web site due to some http links that trigger a warning which is treated as error in the docs auto update.
To look at the latest docs for this release (where there could still be some details that I have not fully updated yet). Please clone the odu/phy repo and then run locally in your pc:
sphinx-build -bhtml -n ./docs ./docs_build/html
Make sure that if you run behind a firewall you set up your proxies before running the above command.
The documentation will be in the path_to_phy/phy/docs/docs_build/html folder and you can open in a browser the index.html.
The corresponding l1 binaries and phycfg and xrancfg files are available at https://github.com/intel/FlexRAN.
Please let me know of any issues as I'm in the process of cleaning up the supporting docs for accuracy.
Thanks
Luis
Luis Farias
The sphinx-build command needs to be issued at the phy folder level
Sindhu Pitta
Hi Luis Farias,
As communicated through the mail, we were able to build all the binaries and run them successfully except FAPI(we are able to build this successfully but while running we can see errors).
we are seeing the below error:
Can you please help us in resolving this?
Thank you
Regards,
Sindhu Pitta.
Luis Farias
Sindhu:
The reason for the problem is that you are still using the old FlexRAN L1 code which is built using DPDK 18.11.(I.e. the bronze release version). Please go to the github.com\intel\FlexRAN and download the latest version which is the oran_e_maintenance_release_v1.0. Same applies to the testmac binary. Keep in mind that we are using now git lfs for the large binary files, so you will need to install it for your local repo and after cloning the FlexRAN repo you will need to do a git lfs pull to retrieve the large binary files otherwise you only get a link to the lfs repo.
Regards
Luis
Sindhu Pitta
Hi Luis,
Thank you for your suggestion.
We are able to solve that primary and secondary version issue by installing dpdk.stable.20.11.1 version and executed all the binaries.
Regards,
Sindhu.
Sindhu Pitta
Hi Luis Farias,
We have cloned the updated code only i.e., oran_e_maintenance_release_v1.0. this one only.
And we can also see the new change sand XML files which have been added wrt radio mode also.
Regards,
Sindhu Pitta
Sindhu Pitta
Hi Harshita Lal,
We are trying pairwise testing between ODU-L and ODU-H.
we are using dpdk.20.11 version for running l1 and fapi binaries.
But for ODU-high we are seeing issues while trying with dpdk.20.11.1 can we use dpdk.19.11 instead?
we tried using dpdk 19.11 and tried pairwise, but couldn't see any fapi messages after running fapi and odu h binaries
below is the fapi log:
Attaching the odu logs below:
Can you help us with this?
Thank you.
Regards,
Sindhu Pitta.
Sindhu Pitta
Hi Harshita Lal,
We tried to build ODU-H for pairwise testing
make odu PHY=INTEL_L1 MACHINE=BIT64 MODE=FDD
here we can see linking errors
we are using dpdk.20.11.1 version as mentioned in the prerequisites for building L1 binaries.
so in the dpdk, we couldn't see .a files but in the makefile, it is linking *.a files.
error:
Can you help us with this?
Thank you.
Regards,
Sindhu Pitta.
Perapalla Maneesha
Hi Luis Farias and Team,
Could you please share the HW specification required to run this new L1 binary? (oran_e_maintenance_release_v1.0)
Thanks in advance.
Regards,
Perapalla Maneesha.
Luis Farias
Hi Perapalla:
HW reqs are the same as the previous versions for the basic test cases provided as examples in timer mode. I will be updating the documentation later this week to provide more details.
Sindhu Pitta
Hi Luis Farias/ Harshita Lal,
We are trying pairwise testing with new e-rel binaries, but seeing errors while compiling odu with l1.
In the new release, the WLS_Open definition has 4 arguments where as in odu high it is using only 3 arguments.
below are the two def of WLS_Open:
in PHY,
#define WLS_MAC_MEMORY_SIZE 0x3EA80000
#define WLS_PHY_MEMORY_SIZE 0x18000000
in odu-h,
#define WLS_MEM_SIZE 0x7ec00000 /* Total size of WLS memory configured *
What should be done to solve this?
Regards,
Sindhu.
Harshita Lal
Hi Sindhu
ODU-High has not been tested yet will latest FlexRAN L1 after the upgrade of L1 to current version. Hence ODU-High and FlexRAN L1 may be out of sync currently.
We will be taking up this activity as part of Integration testing which is yet to be planned based on the availability of test lab setup.
Regards
Harshita.
Sindhu Pitta
Hi Harshita,
Thank you for sharing the info.
We tried building ODU-H with a new L1(new e rel), which uses dpdk 20.11.1.
We have updated the make file of ODU-H also to the same dpdk version i.e., 20.11.1.
But seeing these below linking errors:
ld: cannot find -ldpdk
ld: cannot find -lrte_pmd_thunderx_nicvf
ld: cannot find -lrte_pmd_lio
ld: cannot find -lrte_pmd_vmxnet3_uio
ld: cannot find -lrte_pmd_octeontx_ssovf
ld: cannot find -lrte_pmd_octeontx_zip
make: *** [link_du] Error
And there are no .a files available for the above libraries in the DPDK.20.11.1 version.
Can you suggest any solution to solve these linking errors and work with the new L1 rel?
Is there any code update planned in ODU-H to test with new L1 binaries, If so when can we expect it?
Regards,
Sindhu.
Rajisha Payyanur
Hi Team,
Can anyone please share the details on procuring the smartNIC (mclaren smartNIC)?
Thank you.
Regards,
Rajisha
Perapalla Maneesha
Hi Luis Farias and Team,
We are trying to launch new L1 (i.e oran_e_maintenance_release_v1.0) in radio mode and seeing some memory issues as mentioned below.
Not enough memory available on socket 0! Requested: 20480MB, available: 10240MB
Could anyone please share the server specifications (including /proc/cmdline and /proc/meminfo output) which you are using to test L1 in Radio mode.
Thanks in advance.
Regards,
Perapalla Maneesha.
Perapalla Maneesha
Hi Luis Farias,
We are still not able to solve the memory issue that is occurring when we lauch L1 in radio mode. We modified hugepages to 80 and launched L1 successfully but when we launch L1 with testmac we see another issue in fapi as below :
[ERROR_LOG][SLOT.indication] Unable to create list element. Out of memory!!!
[ERROR_LOG][FAPI MSG HDR] Unable to create list element. Out of memory!!!
Please share the server specifications (including /proc/cmdline and /proc/meminfo output) which you are using to test L1 in Radio mode.
Thanks& Regards,
Perapalla Maneesha.
Luis Farias
Hi Perapalla:
I will share the server specifications that were used for the US O-RAN Plugfest last November. In the plugfest we used two different O-RU simulator vendors and two different L2+ stack vendors for the tests and we were able to perform end to end testing in both cases. If you can share the details on the radio that you are using and it is compliance with the O-RAN CUS Interface it can help to figure out what is wrong. Another aspect is the particular CPU that you are using and the core assignments, you may need to tweak these assignments per the phycfg_xran.xml BbuPoolConfig section, in the testmac_cfg.xml Threads section (or the corresponding configuration for your L2+ app) and in the oran_5g_fapi.cfg MAC2PHY_WORKER, PHY2MAC_WORKER and URLLC_WORKER sections to make sure that they fit within the limitations of the cpu that you are using.
cat /proc/meminfo
Vendor 1 Case Vendor 2 Case
MemTotal: 131602144 kB 131568736 kB
MemFree: 73391424 kB 9944312 kB
MemAvailable: 83114944 kB 45585216 kB
Buffers: 2308 kB 1304 kB
Cached: 11750784 kB 34172036 kB
SwapCached: 0 kB 0 kB
Active: 2136140 kB 25652996 kB
Inactive: 9665832 kB 18088204 kB
Active(anon): 1512024 kB 12403100 kB
Inactive(anon): 810216 kB 1382072 kB
Active(file): 624116 kB 13248988 kB
Inactive(file): 8855616 kB 16706132 kB
Unevictable: 0 kB 408 kB
Mlocked: 0 kB 420 kB
SwapTotal: 16449532 kB 4194300 kB
SwapFree: 16449532 kB 4194300 kB
Dirty: 48 kB 0 kB
Writeback: 0 kB 0 kB
AnonPages: 49092 kB 9567032 kB
Mapped: 25412 kB 398120 kB
Shmem: 2273244 kB 4218152 kB
Slab: 882236 kB 6333968 kB
SReclaimable: 755088 kB 6096232 kB
SUnreclaim: 127148 kB 237736 kB
KernelStack: 18544 kB 51568 kB
PageTables: 5936 kB 296312 kB
NFS_Unstable: 0 kB 0 kB
Bounce: 0 kB 0 kB
WritebackTmp: 0 kB 0 kB
CommitLimit: 61279084 kB 38521388 kB
Committed_AS: 2516216 kB 19807896 kB
VmallocTotal: 34359738367 kB 34359738367 kB
VmallocUsed: 795196 kB 1036884 kB
VmallocChunk: 34358704124 kB 34358247420 kB
Percpu: 12800 kB 16576 kB
HardwareCorrupted: 0 kB 0 kB
CmaTotal: 0 kB 0 kB
CmaFree: 0 kB 0 kB
HugePages_Total: 40 60
HugePages_Free: 22 46
HugePages_Rsvd: 0 0
HugePages_Surp: 0 0
Hugepagesize: 1048576 kB 1048576 kB
DirectMap4k: 740752 kB 143940 kB
DirectMap2M: 10473472 kB 7898944 kB
DirectMap1G: 124780544 kB 127926272 kB
cat /proc/cmdline
Vendor 1
BOOT_IMAGE=/vmlinuz-3.10.0-1127.19.1.rt56.1116.el7.x86_64 root=/dev/mapper/centos-root ro crashkernel=auto
rd.lvm.lv=centos/root rd.lvm.lv=centos/swap rhgb quiet intel_iommu=on iommu=pt usbcore.autosuspend=-1
selinux=0 enforcing=0 nmi_watchdog=0 softlockup_panic=0 audit=0 cgroup_disable=memory mce=off hugepagesz=1G
hugepages=40 hugepagesz=2M hugepages=0 default_hugepagesz=1G irqaffinity=0,29-32,61-63
Vendor 2
BOOT_IMAGE=/vmlinuz-3.10.0-1127.19.1.rt56.1116.el7.x86_64 root=/dev/mapper/centos-root ro crashkernel=auto
rd.lvm.lv=centos/root rd.lvm.lv=centos/swap rhgb quiet LANG=en_US.UTF-8 intel_iommu=on iommu=pt usbcore.autosuspend=-1
selinux=0 enforcing=0 nmi_watchdog=0 softlockup_panic=0 audit=0 intel_pstate=disable cgroup_memory=1 cgroup_enable=memory
mce=off idle=poll hugepagesz=1G hugepages=60 default_hugepagesz=1G processor.max_cstate=1 intel_idle.max_cstate=0
isolcpus=1-27,29-55 rcu_nocbs=1-27,29-55 kthread_cpus=0,28 irqaffinity=0,28 nohz_full=1-27,29-55
Regards
Luis
Perapalla Maneesha
Hi Luis Farias and Harshita Lal,
Thanks for sharing the info.
We are using Benetel RAN 550 as our radio which is compliant with the O-RAN fronthaul interface. But due to some technical issues we didn't launch L1 in RRU presence. Is it mandatory to have RRU in place for launching L1 in radio mode?
We are currently using 6154 processor for our DU which has 4 sockets with 18 cores per socket. We updated our grub as per your /proc/cmdline output with hugepages =120 (i.e 30 per socket) and launched L1 successfully.
But when we run L1 in radio mode, FAPI and testmac in radio mode(phystart 4 10 0) we see below error log in FAPI after certain slot indication messages:
[DEBUG_LOG][FAPI MSG HDR] FAPI Message Header Added for PHY: 0
[DEBUG_LOG][VENDOR EXT indication] No vendor elementfor phy_id=0 yet. Creating new
[DEBUG_LOG][SLOT.indication][0][20,5,0] is_urllc 0
[TRACE_LOG][FAPI2MAC WLS][PUT] 200035270656 size: 44 type: 1a flags: 8200
[TRACE_LOG][FAPI2MAC WLS][PUT] 200035008512 size: 52 type: 82 flags: 8000
[TRACE_LOG][FAPI2MAC WLS][PUT] 200037367808 size: 34136 type: 17 flags: 8100
[TRACE_LOG][FAPI2PHY WLS][GET] 0x2e58680000 size: 48 type: 10 flags: 0
[DEBUG_LOG]To Free 1
[DEBUG_LOG]Free 1
[TRACE_LOG][PHY2MAC] nr5g_fapi_phy2mac_api_recv_handler:
[DEBUG_LOG][FAPI MSG HDR] FAPI Message Header Added for PHY: 0
[DEBUG_LOG][VENDOR EXT indication] No vendor elementfor phy_id=0 yet. Creating new
[ERROR_LOG][VENDOR EXT indication] Unable to create list element. Out of memory!!!
[DEBUG_LOG][SLOT.indication][0][20,6,0] is_urllc 0
[TRACE_LOG][FAPI2MAC WLS][PUT] 200037105664 size: 44 type: 1a flags: 8200
[TRACE_LOG][FAPI2MAC WLS][PUT] 200036843520 size: 52 type: 82 flags: 8100
[TRACE_LOG][FAPI2PHY WLS][GET] 0x2e586c0000 size: 48 type: 10 flags: 0
[DEBUG_LOG]To Free 1
[DEBUG_LOG]Free 1
[TRACE_LOG][PHY2MAC] nr5g_fapi_phy2mac_api_recv_handler:
[ERROR_LOG][FAPI MSG HDR] Unable to create list element. Out of memory!!!
[ERROR_LOG][SLOT.indication] Unable to create list element. Out of memory!!!
[TRACE_LOG][FAPI2PHY WLS][GET] 0x2e58700000 size: 48 type: 10 flags: 0
[DEBUG_LOG]To Free 1
Can you please comment on the above issue and confirm whether RRU is mandatory for L1 testing in radio mode.
Thanks& Regards,
Perapalla Maneesha.
Luis Farias
Hi Perapalla:
A few observations relative to your setup and traces. The testmac that we provide through the github FlexRan binary blob is intended to be used in non-real time mode (i.e. timer mode) using iq samples read from a file into the receiver's memory for UL cases or written by the DL process into the UL memory for the FD cases, so I'm not sure how you run the testmac in xran mode. If I issue from the testmac the phystart 4 10 0, I simply see in the testmac console a print stating "testmac_set_phy_start: mode[4], period[10], count[0]". There is no activity in the FAPI TM or in the L1 consoles as no processing actually takes place since the testmac doesn't have any data to send downlink. Uplink there will be nothing coming as there is no radio and nothing attached remotely as a UE to send PRACH and go through the attachment phase.
All of the end to end tests at the Plugfest used a partner's L2+ stack, although those partners have not contributed yet to the O-RAN OSC an integrated version of their stack with the OSC version of the odu-phy.
In order to get the radio oran mode going it is necessary that in your setup ti run linux PTP to synchronize the Xeon's local time of day. I will attach later traces for L1, 5g FAPI and testmac when starting the L1 in oran radio mode from the l1 shell script i.e. ./l1.sh -xran.
Luis Farias
Hi Perapalla:
Also in the orancfg_sub6.xml you do need to have the appropriate nic card addresses being used in your system and the odu-high stack compatible with this release (Check if RSYS OSC has integrated already with this release as in the ORAN-Plugfest we used the commercial version of the RSYS stack). I will mail you the logs for each one of the consoles when doing ./l1.sh -xran and setting from the testmac the phystart 4 0 10.
Regards
Luis
Perapalla Maneesha
Hi Luis Farias,
Thank a lot for sharing the logs and info.
Our Xeon processor is synchronized with Linux PTP. But, the available odu-h stack i.e. RSYS OSC is still not compatible with the new phy i.e. oran_e_maintenance_release_v1.0.
We are waiting for the new release of odu-h stack which is apparently planned in mid June.
Thanks& Regards,
Perapalla Maneesha.
Rajisha Payyanur
Hi Harshita Lal Manasi Padhy,
We got to know that ‘ODU-High code update as per latest L1 release’ is planned during the defect management sprint ( 6/June/2022 to 24/June/2022).
Is there any further update on this?
We would also like to collaborate for this DU-H enhancement.
Thank you.
Regards,
Rajisha
Rajisha Payyanur
Hi Team,
Any update on this?
Please let us know if any testing support required.
Thank you.
Perapalla Maneesha
Hi Luis Farias and Team,
We have cloned Phy (oran_e_maintenance_release_v1.0) and launched Sample-app with all required vlan settings. But when we capture ecipri packet on RU we could see that vlan tag is missing in packet.
Does FHI has support to append vlan tag while transmitting the packets? As we dig deep into code we see vlan is not added in ethernet header field
Is there any other config required explicitly to enable vlan tags on packets?
We would also like to attach the VLAN field for the PTP packets. Please suggest how to configure it.
Thanks in advance.
Perapalla Maneesha
Hi Luis Farias, Harshita Lal and Team,
We are facing certain issues when we are trying to integrate DU with third party RRU.
Please help us on how we can integrate ODU-H and ODU-L with third party RRU's.
Regards,
Perapalla Maneesha
hoejoo lee
Does anyone know how to add project page link in the drop-down list at the "Projects" tab ??
I want to add aimlfw project link ~ please let me know the person in charge of it
GIANLUCA CAMILLO VALASTRO
Dear all,
We are trying to run RU and DU on the same instance. We deployed the sample application on AWS EC2 C5 instance. We have ENA interfaces, with PCIe addresses like these 0000:00:06.0, 0000:00:07.0, etc. We configured .cfg and .dat files, the RU and DU start correctly, but the output shows only tx and no RX traffic (the related counters in output are all zeros) for both RU and DU.
Any suggestion to solve this no RX traffic issue? Anyone of you is able to run correctly the sample application on AWS EC2 instances?
Thank you for your availability and for your support.
Best regards
Gianluca Camillo Valastro
Le On
Hello all,
we are currently in the process of setting up a (ideally single-node) Open-RAN test environment. The test environment should have simulators, which simulate for example the O-D, O-RU and UE. The goal of the test environment is to obtain a fully virtualized O-RAN (non and near-RTRIC).
Currently, I am facing the problem that the deployment of the individual components is quite simple, but I cannot find a way to get a near-RTRIC to communicate with a non-RTRIC.
Does anyone have a tip for me on the easiest way to set up a complete O-RAN environment?
Has anyone done something similar in the past?
Best Regards
Leon
John Keeney
Hi Le On
Please post this question to the Mailing list for tech support - and please include details of your environment and the full logs.
(OSC "discuss" mailing list: https://lists.o-ran-sc.org/g/discuss: discuss@lists.o-ran-sc.org Please use hashtags #nonrtric and #ricp)
In the email perhaps you can give more detail about what is going wrong and what you have already tried?
Comments on wiki pages are not very suitable support requests.
Thanks,
John & team