Hi, Remi, I just have one dispatcher, both devices linked to this dispatcher. And the issue continues after change to next:
{ "port": 3079, "blocksize": 4096, "poll_delay": 3, "coordinator_hostname": "10.191.253.109" }
On Thu, Jul 1, 2021 at 4:55 PM Remi Duraffort remi.duraffort@linaro.org wrote:
Hello,
by default, lava-coordinator is expected to run on localhost. So the second device will connect to localhost instead of the real host hosting coordinator. Change the host into /etc/lava-coordinator/lava-coordinator.conf
Le mar. 29 juin 2021 à 07:19, Hedy Lamarr lamarrhedy97@gmail.com a écrit :
The "ssh" device will ssh to another machine which is not on the same dispatcher to start a iperf server. The "dragonboard-410c" device will start a docker container, then in this container, it will call iperf client to connect to the iperf server.
- In lava admin page, I link both "ssh device" & " dragonboard-410c" to
the same worker. 2. But, the command in ssh(iperf server) will run on another machine, while command in docker container(iperf client) run on the same machine of worker I think. I'm not sure you mean 1 or 2?
On Mon, Jun 28, 2021 at 10:34 PM Remi Duraffort < remi.duraffort@linaro.org> wrote:
Hello,
Le lun. 28 juin 2021 à 08:26, Hedy Lamarr lamarrhedy97@gmail.com a écrit :
Hello,
What additional I need to afford to debug this issue?
Thanks, Hedy Lamarr
On Thu, Jun 17, 2021 at 4:34 PM Hedy Lamarr lamarrhedy97@gmail.com wrote:
YES, to make it clear, I restart the lava server just now and give you a full log when that multinode job run:
2021-06-17 09:16:07,428 INFO [INIT] LAVA coordinator has started. 2021-06-17 09:16:07,757 INFO [INIT] Version 2021.03 2021-06-17 09:16:07,757 INFO [INIT] Loading configuration from /etc/lava-coordinator/lava-coordinator.conf 2021-06-17 09:16:08,076 INFO [BTSP] binding to 0.0.0.0:3079 2021-06-17 09:16:08,076 INFO Ready to accept new connections 2021-06-17 09:17:23,603 INFO The decbbfe5-f3be-4e6c-a2b8-5744eabfe8a7 group will contain 2 nodes. 2021-06-17 09:17:23,603 INFO Waiting for 1 more clients to connect to decbbfe5-f3be-4e6c-a2b8-5744eabfe8a7 group 2021-06-17 09:17:23,603 INFO Ready to accept new connections 2021-06-17 09:17:23,790 INFO Group complete, starting tests 2021-06-17 09:17:23,790 INFO Ready to accept new connections 2021-06-17 09:17:26,613 INFO Group complete, starting tests 2021-06-17 09:17:26,613 INFO Ready to accept new connections 2021-06-17 09:18:03,522 DEBUG clear Group Data: 1 of 2 2021-06-17 09:18:03,522 INFO Ready to accept new connections 2021-06-17 09:18:06,001 DEBUG clear Group Data: 2 of 2 2021-06-17 09:18:06,001 DEBUG Clearing group data for decbbfe5-f3be-4e6c-a2b8-5744eabfe8a7 2021-06-17 09:18:06,001 INFO Ready to accept new connections 2021-06-17 09:24:43,620 INFO The 8956d8e7-1097-43e0-95dd-7afc61b2908b group will contain 2 nodes. 2021-06-17 09:24:43,620 INFO Waiting for 1 more clients to connect to 8956d8e7-1097-43e0-95dd-7afc61b2908b group 2021-06-17 09:24:43,620 INFO Ready to accept new connections 2021-06-17 09:24:43,871 INFO Group complete, starting tests 2021-06-17 09:24:43,871 INFO Ready to accept new connections 2021-06-17 09:24:46,634 INFO Group complete, starting tests 2021-06-17 09:24:46,634 INFO Ready to accept new connections 2021-06-17 09:25:45,746 INFO lava_send: {'port': 3079, 'blocksize': 4096, 'poll_delay': 3, 'host': '10.191.253.109', 'hostname': 'lavaslave1', 'client_name': '3077', 'group_name': '8956d8e7-1097-43e0-95dd-7afc61b2908b', 'role': 'host', 'request': 'lava_send', 'messageID': 'server_ready', 'message': {}} 2021-06-17 09:25:45,747 INFO lavaSend handler in Coordinator received a messageID 'server_ready' for group '8956d8e7-1097-43e0-95dd-7afc61b2908b' from 3077 2021-06-17 09:25:45,747 DEBUG message ID server_ready {"3077": {}} for 3077 2021-06-17 09:25:45,747 DEBUG broadcast ID server_ready {"3077": {}} for 3076 2021-06-17 09:25:45,747 DEBUG broadcast ID server_ready {"3077": {}} for 3077 2021-06-17 09:25:45,747 INFO Ready to accept new connections
This log similar to the log I saw on web, the "SSH device" with lava_send looks ok, but "dragonboard device for android test" with lava_wait looks not ok, it's just hung. From above log, looks the coordinator did not receive anything?
For what I see in the logs, lava-coordinator is not receiving any signal from the second test.
Are both devices on the same dispatcher/worker?
On Thu, Jun 17, 2021 at 4:02 PM Remi Duraffort <
remi.duraffort@linaro.org> wrote:
Le jeu. 17 juin 2021 à 09:11, Hedy Lamarr lamarrhedy97@gmail.com a écrit :
> The output is: > > service lava-coordinator status > ● lava-coordinator.service - LAVA coordinator > Loaded: loaded (/lib/systemd/system/lava-coordinator.service; > enabled; vendor preset: enabled) > Active: active (running) since Fri 2021-06-04 18:09:19 CET; 1 > weeks 5 days ago > Main PID: 629 (lava-coordinato) > Tasks: 1 (limit: 4915) > Memory: 7.4M > CGroup: /system.slice/lava-coordinator.service > └─629 /usr/bin/python3 /usr/bin/lava-coordinator > --loglevel DEBUG >
So it's working.
Is it listening on 10.191.253.109:3079 ? Do you have anything in the lava-coordinator logs? (/var/log/lava-coordinator.log)
> On Thu, Jun 17, 2021 at 3:05 PM Remi Duraffort < > remi.duraffort@linaro.org> wrote: > >> >> >> Le jeu. 17 juin 2021 à 09:02, Hedy Lamarr lamarrhedy97@gmail.com >> a écrit : >> >>> Hello Remi, >>> >>> I think lava-coordinator is running. >>> >>> Because there are 2 devices here: >>> Device1: dragonboard-410c, when lava-wait server_ready, it hangs >>> with above log. >>> Device2: ssh, when lava-send server_ready, it shows: Connecting to >>> LAVA Coordinator on 10.191.253.109:3079 timeout=300 seconds. >>> >>> Would it be possible that lava-coordinator just works for ssh, but >>> not for dragonboard-410c? >>> Also I think the netstat, it shows: >>> tcp 0 0 0.0.0.0:3079 0.0.0.0:* >>> LISTEN 629/python3 off (0.00/0/0) >>> Does this mean coordinator running? Or how can I make sure >>> coordinator running? >>> >> >> service lava-coordinator status >> >> >>> >>> Thanks, >>> Hedy Lamarr >>> >>> On Thu, Jun 17, 2021 at 2:33 PM Remi Duraffort < >>> remi.duraffort@linaro.org> wrote: >>> >>>> Hello, >>>> >>>> do you have lava-coordinator running? >>>> >>>> Le lun. 14 juin 2021 à 14:29, Hedy Lamarr lamarrhedy97@gmail.com >>>> a écrit : >>>> >>>>> By the way, we use 2021.03.post1. >>>>> >>>>> On Wed, Jun 9, 2021 at 10:40 AM Hedy Lamarr < >>>>> lamarrhedy97@gmail.com> wrote: >>>>> >>>>>> Dear community, >>>>>> >>>>>> We are new to lava and try to use lava in our android test. We >>>>>> have issues when test iperf. >>>>>> >>>>>> Job: >>>>>> >>>>>> job_name: android iperf test >>>>>> timeouts: >>>>>> job: >>>>>> minutes: 10080 >>>>>> action: >>>>>> minutes: 120 >>>>>> connection: >>>>>> minutes: 5 >>>>>> priority: medium >>>>>> visibility: public >>>>>> protocols: >>>>>> lava-multinode: >>>>>> roles: >>>>>> device: >>>>>> count: 1 >>>>>> device_type: dragonboard-410c >>>>>> timeout: >>>>>> minutes: 5 >>>>>> host: >>>>>> count: 1 >>>>>> device_type: ssh >>>>>> timeout: >>>>>> minutes: 5 >>>>>> context: >>>>>> ssh_host: localhost >>>>>> ssh_user: root >>>>>> ssh_port: 22 >>>>>> ssh_identity_file: /root/.ssh/id_rsa >>>>>> actions: >>>>>> - deploy: >>>>>> role: >>>>>> - host >>>>>> timeout: >>>>>> minutes: 2 >>>>>> to: ssh >>>>>> os: debian >>>>>> - boot: >>>>>> role: >>>>>> - host >>>>>> method: ssh >>>>>> connection: ssh >>>>>> prompts: >>>>>> - '@labpc1' >>>>>> - test: >>>>>> role: >>>>>> - host >>>>>> timeout: >>>>>> minutes: 120 >>>>>> definitions: >>>>>> - from: inline >>>>>> name: smoke-case >>>>>> path: inline/test.yaml >>>>>> repository: >>>>>> metadata: >>>>>> format: Lava-Test Test Definition >>>>>> name: smoke >>>>>> description: Run smoke case >>>>>> run: >>>>>> steps: >>>>>> - sleep 60 >>>>>> - lava-send "server_ready" >>>>>> - iperf -s -V -P 1 >>>>>> - test: >>>>>> role: >>>>>> - device >>>>>> definitions: >>>>>> - from: inline >>>>>> name: cts_cts-media_test >>>>>> path: inline/cts_cts-media_test.yaml >>>>>> repository: >>>>>> metadata: >>>>>> description: cts cts-media test run >>>>>> format: Lava-Test Test Definition 1.0 >>>>>> name: cts-cts-media-test-run >>>>>> run: >>>>>> steps: >>>>>> - adb wait-for-device >>>>>> - adb devices >>>>>> - adb root >>>>>> - adb wait-for-device >>>>>> - adb devices >>>>>> - lava-wait "server_ready" >>>>>> - sleep 3 >>>>>> - lava-test-case "Case1" --shell adb shell >>>>>> /data/local/iperf -c 10.191.253.21 -t 10 >>>>>> docker: >>>>>> image: terceiro/android-platform-tools >>>>>> timeout: >>>>>> minutes: 4200 >>>>>> >>>>>> The job log for dragonboard-410c is: >>>>>> + lava-wait server_ready >>>>>> <LAVA_WAIT_DEBUG preparing Wed Jun 8 10:07:22 CST 2021> >>>>>> <LAVA_WAIT_DEBUG started Wed Jun 8 10:07:22 CST 2021> >>>>>> <LAVA_MULTI_NODE> <LAVA_WAIT server_ready> >>>>>> <LAVA_WAIT_DEBUG finished Wed Jun 8 10:07:22 CST 2021> >>>>>> <LAVA_WAIT_DEBUG finished Wed Jun 8 10:07:22 CST 2021> >>>>>> <LAVA_WAIT_DEBUG starting to wait Wed Jun 8 10:07:22 CST 2021> >>>>>> NOTE: it looks hung at this step, the job can't continue. >>>>>> >>>>>> The job log for ssh is: >>>>>> + lava-send server_ready >>>>>> <LAVA_SEND_DEBUG lava_multi_node_send preparing Wed Jun 8 >>>>>> 10:07:53 CST 2021> >>>>>> <LAVA_SEND_DEBUG _lava_multi_node_send started Wed Jun 8 >>>>>> 10:07:53 CST 2021> >>>>>> <LAVA_MULTI_NODE> <LAVA_SEND server_ready> >>>>>> Received Multi_Node API <LAVA_SEND> >>>>>> messageID: SEND-server_ready >>>>>> lava-multinode lava-send >>>>>> Handling signal <LAVA_SEND {"request": "lava_send", >>>>>> "messageID": "server_ready", "message": {}, "timeout": 300}> >>>>>> Setting poll timeout of 300 seconds >>>>>> requesting lava_send server_ready >>>>>> message: {} >>>>>> requesting lava_send server_ready with args {} >>>>>> request_send server_ready {} >>>>>> Sending {'request': 'lava_send', 'messageID': 'server_ready', >>>>>> 'message': {}} >>>>>> final message: {"port": 3079, "blocksize": 4096, "poll_delay": >>>>>> 3, "host": "10.191.253.109", "hostname": "lavaslave1", "client_name": >>>>>> "3035", "group_name": "8a362e2a-6ee9-4f48-bddb-378ac2425f06", "role": >>>>>> "host", "request": "lava_send", "messageID": "server_ready", "message": {}} >>>>>> Connecting to LAVA Coordinator on 10.191.253.109:3079 >>>>>> timeout=300 seconds. >>>>>> case: multinode-send-server_ready >>>>>> case_id: 39177 >>>>>> definition: 0_smoke-case >>>>>> result: pass >>>>>> <LAVA_SEND_DEBUG _lava_multi_node_send finished Wed Jun 8 >>>>>> 10:07:53 CST 2021> >>>>>> <LAVA_SEND_DEBUG lava_multi_node_send finished Wed Jun 8 >>>>>> 10:07:53 CST 2021> >>>>>> + iperf -s -V -P 1 >>>>>> ------------------------------------------------------------ >>>>>> Server listening on TCP port 5001 >>>>>> TCP window size: 85.3 KByte (default) >>>>>> ------------------------------------------------------------ >>>>>> >>>>>> It seems that one node can't receive server-ready from another >>>>>> node, what's wrong with my job? Please help! >>>>>> >>>>>> Thanks, >>>>>> Hedy Lamarr >>>>>> >>>>>> _______________________________________________ >>>>> Lava-users mailing list >>>>> Lava-users@lists.lavasoftware.org >>>>> https://lists.lavasoftware.org/mailman/listinfo/lava-users >>>>> >>>> >>>> >>>> -- >>>> Rémi Duraffort >>>> TuxArchitect >>>> Linaro >>>> >>> >> >> -- >> Rémi Duraffort >> TuxArchitect >> Linaro >> >
-- Rémi Duraffort TuxArchitect Linaro
-- Rémi Duraffort TuxArchitect Linaro
-- Rémi Duraffort TuxArchitect Linaro