search

shanshanpt / seastar_example

seastar example
8 stars 6 forks source link

run server with dpdk failed #4

Open xiaqunfeng opened 5 years ago

xiaqunfeng commented 5 years ago

I successfully built server and client. When I run server with posix satck, it running successful:

# ./server --port=13001 --smp=1 --cpuset=5
WARN  2019-07-01 20:48:02,704 [shard 0] seastar - Unable to set SCHED_FIFO scheduling policy for timer thread; latency impact possible. Try adding CAP_SYS_NICE
TCP Echo-Server listen on: 13001
current 0 total 0 qps 0
current 0 total 0 qps 0
current 0 total 0 qps 0
current 0 total 0 qps 0
current 0 total 0 qps 0
current 0 total 0 qps 0
current 0 total 0 qps 0
...
... // waiting for worker
...
current 1 total 1 qps 65054
current 1 total 1 qps 130049
current 1 total 1 qps 135699
current 1 total 1 qps 126766
current 1 total 1 qps 132650
current 1 total 1 qps 127944
current 1 total 1 qps 135250

When I run server with dpdk, it show me the error :

# ./server --port=13001 --smp=1 --cpuset=5 --dpdk-pmd --network-stack=native
EAL: Detected 56 lcore(s)
EAL: Detected 2 NUMA nodes
EAL: Multi-process socket /var/run/dpdk/rte/mp_socket
EAL: No available hugepages reported in hugepages-2048kB
EAL: FATAL: Cannot get hugepage information.
EAL: Cannot get hugepage information.
EAL: Error - exiting with code: 1
  Cause: Cannot init EAL

Could you give us any help?

shanshanpt commented 5 years ago

I successfully built server and client. When I run server with posix satck, it running successful:

# ./server --port=13001 --smp=1 --cpuset=5
WARN  2019-07-01 20:48:02,704 [shard 0] seastar - Unable to set SCHED_FIFO scheduling policy for timer thread; latency impact possible. Try adding CAP_SYS_NICE
TCP Echo-Server listen on: 13001
current 0 total 0 qps 0
current 0 total 0 qps 0
current 0 total 0 qps 0
current 0 total 0 qps 0
current 0 total 0 qps 0
current 0 total 0 qps 0
current 0 total 0 qps 0
...
... // waiting for worker
...
current 1 total 1 qps 65054
current 1 total 1 qps 130049
current 1 total 1 qps 135699
current 1 total 1 qps 126766
current 1 total 1 qps 132650
current 1 total 1 qps 127944
current 1 total 1 qps 135250

When I run server with dpdk, it show me the error :

# ./server --port=13001 --smp=1 --cpuset=5 --dpdk-pmd --network-stack=native
EAL: Detected 56 lcore(s)
EAL: Detected 2 NUMA nodes
EAL: Multi-process socket /var/run/dpdk/rte/mp_socket
EAL: No available hugepages reported in hugepages-2048kB
EAL: FATAL: Cannot get hugepage information.
EAL: Cannot get hugepage information.
EAL: Error - exiting with code: 1
  Cause: Cannot init EAL

Could you give us any help?

You should configure the hugepage firstly. The steps like, as follows

sudo sh ./dpdk-setup.sh
Find "Setup hugepage mappings for NUMA systems"
sudo mkdir /mnt/huge
mount -t hugetlbfs nodev /mnt/huge
...
xiaqunfeng commented 5 years ago

Thank you for your reply. When I choose "Setup hugepage mappings for NUMA systems", it always tells me permission denied. The output as follows:

# sudo bash ./dpdk/usertools/dpdk-setup.sh
...
[45] Setup hugepage mappings for NUMA systems
...
Option: 45

Removing currently reserved hugepages
.echo_tmp: 2: .echo_tmp: cannot create /sys/devices/system/node/node0/hugepages/hugepages-2048kB/nr_hugepages: Read-only file system
.echo_tmp: 3: .echo_tmp: cannot create /sys/devices/system/node/node1/hugepages/hugepages-2048kB/nr_hugepages: Read-only file system
Unmounting /mnt/huge and removing directory

  Input the number of 2048kB hugepages for each node
  Example: to have 128MB of hugepages available per node in a 2MB huge page system,
  enter '64' to reserve 64 * 2MB pages on each node
Number of pages for node0: 64
Number of pages for node1: 64
Reserving hugepages
.echo_tmp: 2: .echo_tmp: cannot create /sys/devices/system/node/node0/hugepages/hugepages-2048kB/nr_hugepages: Read-only file system
.echo_tmp: 3: .echo_tmp: cannot create /sys/devices/system/node/node1/hugepages/hugepages-2048kB/nr_hugepages: Read-only file system
Creating /mnt/huge and mounting as hugetlbfs
mount: permission denied

Press enter to continue ...
shanshanpt commented 5 years ago

Thank you for your reply. When I choose "Setup hugepage mappings for NUMA systems", it always tells me permission denied. The output as follows:

# sudo bash ./dpdk/usertools/dpdk-setup.sh
...
[45] Setup hugepage mappings for NUMA systems
...
Option: 45

Removing currently reserved hugepages
.echo_tmp: 2: .echo_tmp: cannot create /sys/devices/system/node/node0/hugepages/hugepages-2048kB/nr_hugepages: Read-only file system
.echo_tmp: 3: .echo_tmp: cannot create /sys/devices/system/node/node1/hugepages/hugepages-2048kB/nr_hugepages: Read-only file system
Unmounting /mnt/huge and removing directory

  Input the number of 2048kB hugepages for each node
  Example: to have 128MB of hugepages available per node in a 2MB huge page system,
  enter '64' to reserve 64 * 2MB pages on each node
Number of pages for node0: 64
Number of pages for node1: 64
Reserving hugepages
.echo_tmp: 2: .echo_tmp: cannot create /sys/devices/system/node/node0/hugepages/hugepages-2048kB/nr_hugepages: Read-only file system
.echo_tmp: 3: .echo_tmp: cannot create /sys/devices/system/node/node1/hugepages/hugepages-2048kB/nr_hugepages: Read-only file system
Creating /mnt/huge and mounting as hugetlbfs
mount: permission denied

Press enter to continue ...

emmm... If you are in docker environment, use --privileged. In host environment, use sudo ?