Closed y-koj closed 1 year ago
And here is the full output of active standby. I could not include it in the first comment because it met the character count limit.
It seems that your user does not have the right permission, could you modify the script to include 'sudo' or use a root user? BTW, you can use sudo lsmod
to see if these kernel modules are properly installed.
Thank you for the quick response! I switched to root and the scripts worked correctly!
Description
Liteflow does not work properly on Ubuntu 18.04 and Ubuntu 20.04.
The following warnings appeared while building kernel modules, which tells that
lf_register_model
is not visible from the liteflow models.Environment
I met this error on Ubuntu 20.04 on the real machine. I also tried on Ubuntu 18.04 on the virtual machine, but it failed with the same warning messages.
Ubuntu 20.04
Ubuntu 18.04
Steps to reproduce
apt update && apt upgrade
because it was a fresh machine.$ cd liteflow
$ sudo apt install -y build-essential libnl-3-dev libnl-genl-3-dev pkg-config linux-headers-generic python3 python3-pip
$ pip3 install -U pip $ pip3 install -r bin/requirements.txt
$ make all $ make module_install $ make tcp_kernel_install
Additional Information
The result of
sysctl net.ipv4.tcp_congestion_control
tells that liteflow is used for congestion control.Full output of make command
``` yk@ubuntu18-on-qemu:~/liteflow$ make all make clean make[1]: Entering directory '/home/yk/liteflow' make[1]: Leaving directory '/home/yk/liteflow' make module make[1]: Entering directory '/home/yk/liteflow' make[2]: Entering directory '/home/yk/liteflow/build/datapath' make -C /lib/modules/4.15.0-210-generic/build M=/home/yk/liteflow/build/datapath modules make[3]: Entering directory '/usr/src/linux-headers-4.15.0-210-generic' CC [M] /home/yk/liteflow/build/datapath/liteflow.o CC [M] /home/yk/liteflow/build/datapath/liteflow_model.o CC [M] /home/yk/liteflow/build/datapath/liteflow_nl.o LD [M] /home/yk/liteflow/build/datapath/lf_kernel.o CC [M] /home/yk/liteflow/build/datapath/liteflow_tcp.o LD [M] /home/yk/liteflow/build/datapath/lf_tcp_kernel.o CC [M] /home/yk/liteflow/build/datapath/liteflow_netfilter.o LD [M] /home/yk/liteflow/build/datapath/lf_netfilter_kernel.o Building modules, stage 2. MODPOST 3 modules CC /home/yk/liteflow/build/datapath/lf_kernel.mod.o LD [M] /home/yk/liteflow/build/datapath/lf_kernel.ko CC /home/yk/liteflow/build/datapath/lf_netfilter_kernel.mod.o LD [M] /home/yk/liteflow/build/datapath/lf_netfilter_kernel.ko CC /home/yk/liteflow/build/datapath/lf_tcp_kernel.mod.o LD [M] /home/yk/liteflow/build/datapath/lf_tcp_kernel.ko make[3]: Leaving directory '/usr/src/linux-headers-4.15.0-210-generic' make[2]: Leaving directory '/home/yk/liteflow/build/datapath' make[1]: Leaving directory '/home/yk/liteflow' make lib make[1]: Entering directory '/home/yk/liteflow' mkdir: cannot create directory ‘build’: File exists Makefile:26: recipe for target 'lib' failed make[1]: [lib] Error 1 (ignored) make[2]: Entering directory '/home/yk/liteflow/build/lib' gcc -c -Werror -O3 -fpic liblf_dp_notification.c -I../include -I/usr/include/libnl3 -lnl-genl-3 -lnl-3 gcc -shared -o liblf.so liblf_dp_notification.o -I/usr/include/libnl3 -lnl-genl-3 -lnl-3 make[2]: Leaving directory '/home/yk/liteflow/build/lib' make[1]: Leaving directory '/home/yk/liteflow' make test make[1]: Entering directory '/home/yk/liteflow' mkdir: cannot create directory ‘build’: File exists Makefile:32: recipe for target 'test' failed make[1]: [test] Error 1 (ignored) make[2]: Entering directory '/home/yk/liteflow/build/test' make[3]: Entering directory '/home/yk/liteflow/build/test/kernel' make -C /lib/modules/4.15.0-210-generic/build M=/home/yk/liteflow/build/test/kernel modules make[4]: Entering directory '/usr/src/linux-headers-4.15.0-210-generic' CC [M] /home/yk/liteflow/build/test/kernel/liteflow_sample_model.o LD [M] /home/yk/liteflow/build/test/kernel/lf_sample_model.o Building modules, stage 2. MODPOST 1 modules WARNING: "lf_unregister_model" [/home/yk/liteflow/build/test/kernel/lf_sample_model.ko] undefined! WARNING: "lf_register_model" [/home/yk/liteflow/build/test/kernel/lf_sample_model.ko] undefined! WARNING: "lf_unregister_app" [/home/yk/liteflow/build/test/kernel/lf_sample_model.ko] undefined! WARNING: "lf_register_app" [/home/yk/liteflow/build/test/kernel/lf_sample_model.ko] undefined! WARNING: "lf_query_model" [/home/yk/liteflow/build/test/kernel/lf_sample_model.ko] undefined! CC /home/yk/liteflow/build/test/kernel/lf_sample_model.mod.o LD [M] /home/yk/liteflow/build/test/kernel/lf_sample_model.ko make[4]: Leaving directory '/usr/src/linux-headers-4.15.0-210-generic' make[3]: Leaving directory '/home/yk/liteflow/build/test/kernel' gcc -Werror -O3 -o test_dp_notification -I../include test_dp_notification.c -L../lib -llf -I/usr/include/libnl3 -lnl-genl-3 -lnl-3 make[2]: Leaving directory '/home/yk/liteflow/build/test' make[1]: Leaving directory '/home/yk/liteflow' ```Full output of snapshot generation
``` yk@ubuntu18-on-qemu:~/liteflow/script$ ./snapshot_generation.sh 2023-05-14 10:56:20.170478: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'libnvinfer.so.6'; dlerror: libnvinfer.so.6: cannot open shared object file: No such file or directory 2023-05-14 10:56:20.170522: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'libnvinfer_plugin.so.6'; dlerror: libnvinfer_plugin.so.6: cannot open shared object file: No such file or directory 2023-05-14 10:56:20.170527: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:30] Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly. 2023-05-14 10:56:20.409170: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'libcuda.so.1'; dlerror: libcuda.so.1: cannot open shared object file: No such file or directory 2023-05-14 10:56:20.409191: E tensorflow/stream_executor/cuda/cuda_driver.cc:351] failed call to cuInit: UNKNOWN ERROR (303) 2023-05-14 10:56:20.409207: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (ubuntu18-on-qemu): /proc/driver/nvidia/version does not exist 2023-05-14 10:56:20.409395: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA 2023-05-14 10:56:20.412070: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2112000000 Hz 2023-05-14 10:56:20.412146: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x4905f60 initialized for platform Host (this does not guarantee that XLA will be used). Devices: 2023-05-14 10:56:20.412154: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): Host, Default Version WARNING:tensorflow:Unable to create a python object for variable