Closed llermaly closed 1 year ago
Not sure if this is related but I see this when running docker before the marqo success
dockerd is running OpenSearch not found; running OpenSearch Unable to find image 'marqoai/marqo-os:0.0.2' locally Error response from daemon: No such container: marqo-os Error: failed to start containers: marqo-os 0.0.2: Pulling from marqoai/marqo-os 017c82d6391f: Pull complete d192a8214bfc: Pull complete e3d96ac441d0: Pull complete 58812b24fd36: Pull complete 4f4fb700ef54: Pull complete cd70a404e73a: Pull complete 0d27a856ae42: Pull complete b44bb45da9a5: Pull complete 1340728d1abc: Pull complete Digest: sha256:f5320b8778f4be248d7affe797898711f9857985c502adc6c4346ca825170897 Status: Downloaded newer image for marqoai/marqo-os:0.0.2 fb3ef73b57532b252546628688cb8d5ce4779fbdf5b845ffe0b3174c2a10ac00 time="2022-10-16T19:15:57.844286896Z" level=info msg="loading plugin \"io.containerd.event.v1.publisher\"..." runtime=io.containerd.runc.v2 type=io.containerd.event.v1 time="2022-10-16T19:15:57.844393329Z" level=info msg="loading plugin \"io.containerd.internal.v1.shutdown\"..." runtime=io.containerd.runc.v2 type=io.containerd.internal.v1 time="2022-10-16T19:15:57.844404971Z" level=info msg="loading plugin \"io.containerd.ttrpc.v1.task\"..." runtime=io.containerd.runc.v2 type=io.containerd.ttrpc.v1 time="2022-10-16T19:15:57.845128064Z" level=info msg="starting signal loop" namespace=moby path=/run/docker/containerd/daemon/io.containerd.runtime.v2.task/moby/fb3ef73b57532b252546628688cb8d5ce4779fbdf5b845ffe0b3174c2a10ac00 pid=1667 runtime=io.containerd.runc.v2 time="2022-10-16T19:15:58.059404481Z" level=error msg="failed to enable controllers ([cpuset cpu io memory hugetlb pids rdma misc])" error="failed to write subtree controllers [cpuset cpu io memory hugetlb pids rdma misc] to \"/sys/fs/cgroup/docker/cgroup.subtree_control\": write /sys/fs/cgroup/docker/cgroup.subtree_control: no such file or directory" time="2022-10-16T19:15:58.061049215Z" level=warning msg="error from *cgroupsv2.Manager.EventChan" error="failed to add inotify watch for \"/sys/fs/cgroup/docker/fb3ef73b57532b252546628688cb8d5ce4779fbdf5b845ffe0b3174c2a10ac00/memory.events\": no such file or directory"
Thanks @llermaly . I will take a look. A couple of questions 1) which commands did you use to start marqo? 2) how long after starting marqo did you try and run the simple-wiki script? the first run takes a little longer as there are a couple of downloads 3) could you run any of the readme examples?
Hi @jn2clark I started with the ones from the main README :
docker rm -f marqo;
docker pull marqoai/marqo:0.0.5;
docker run --name marqo -it --privileged -p 8882:8882 --add-host host.docker.internal:host-gateway marqoai/marqo:0.0.5
Then installing the python client and ran the python script
python3 simple_wiki_demo.py
What I noticed is marqo-os was not running after doing this. Tried to run the marqo-os container manually but no luck, same error.
I will try removing all the containers/images/volumes and trying again
Tried spinning up opensearch independently as the M1 guide says and spins up but once I run the demo py script opensearch crashes :
Killing performance analyzer process 34 ./opensearch-docker-entrypoint.sh: line 116: 33 Killed "$@" "${opensearch_opts[@]}" OpenSearch exited with code 137 Performance analyzer exited with code 143
Are you running on an arm based machine? If not, you will need to modify the command slighlty:
docker rm -f marqo-os; docker run -p 9200:9200 -p 9600:9600 -e "discovery.type=single-node" marqoai/marqo-os:0.0.2
Note the non -arm
tag
Yeah, that's what I did. Opensearch comes up good, I can curl localhost:9200 (admin/admin creds) and see the cluster status. But once I run the python script it goes down
At least running both things separately I can ingest some docs, so the issue is in the python script
Thanks. It could be related to the system RAM. What machine are you running it on? You can try and change this line https://github.com/marqo-ai/marqo/blob/mainline/examples/SimpleWiki/simple_wiki_demo.py#L69 to use only a single process and see if that works.
responses = client.index(index_name).add_documents(data, device=device, processes=1, batch_size=50)
No luck reducing processes. This is the machine:
4 GB Memory / 2 Intel vCPUs / 80 GB Disk
I will try bumping up
Ah, yeah might be worth moving to 8GB RAM
Thank you, it is not crashing but loading now.
Do you have any idea why marqos-os is not spinning up with the initial run command?
Great! Not sure why the first method did not work. We will take a look and see if we can reproduce. My only other suggestion (might be a long-shot) is to upgrade docker and try again.
I had similar issues before with this example, caused by two reasons:
My suggestion would be to make that demo smaller to run on less powerful hardware.
I updated the demo to automatically handle excessively large documents. https://github.com/marqo-ai/marqo/blob/mainline/examples/SimpleWiki/simple_wiki_demo.py#L57 . if you have any other problems, you can re-open!
Describe the bug Server error 500 trying to run the SimpleWiki example. Tried with 0.0.5 and 0.0.3 as the demo shows.
To Reproduce Steps to reproduce the behavior:
Expected behavior Demo should work
Desktop (please complete the following information):