Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Bug]: [benchmark] Failed to create diskann index with error:code=1, message= 450737097756315587: failed to create index, C Runtime Exception: index not supported #34222

Open
1 task done
elstic opened this issue Jun 27, 2024 · 8 comments
Assignees
Labels
kind/bug Issues or changes related a bug test/benchmark benchmark test triage/accepted Indicates an issue or PR is ready to be actively worked on.
Milestone

Comments

@elstic
Copy link
Contributor

elstic commented Jun 27, 2024

Is there an existing issue for this?

  • I have searched the existing issues

Environment

- Milvus version: 2.4-20240626-07a05db5-amd64
- Deployment mode(standalone or cluster): all
- MQ type(rocksmq, pulsar or kafka):    
- SDK version(e.g. pymilvus v2.0.0rc2): pymilvus 2.4.4rc8
- OS(Ubuntu or CentOS): 
- CPU/Memory: 
- GPU: 
- Others:

Current Behavior

  1. deploy milvus
  2. insert data
  3. build diskann index

test env: 4am , qa-milvus ns

fouram-disk-sta24800-6-88-6723-etcd-0                             1/1     Running                           0                4m8s    10.104.25.26    4am-node30   <none>           <none>
fouram-disk-sta24800-6-88-6723-milvus-standalone-ff46bbc9csqxtq   1/1     Running                           3 (2m22s ago)    4m8s    10.104.16.12    4am-node21   <none>           <none>
fouram-disk-sta24800-6-88-6723-minio-5477cd5c9d-4hpzh             1/1     Running                           0                4m8s    10.104.16.11    4am-node21   <none>           <none> (base.py:261)

client error log

[2024-06-26 18:15:02,242 -  INFO - fouram]: [Base] Index params of fouram_jDJJmh2l:[{'float_vector': {'index_type': 'DISKANN', 'metric_type': 'L2', 'params': {}}}] (base.py:486)
[2024-06-26 18:15:02,242 -  INFO - fouram]: [Base] Start build index of DISKANN for field:float_vector collection:fouram_jDJJmh2l, params:{'index_type': 'DISKANN', 'metric_type': 'L2', 'params': {}}, kwargs:{} (base.py:467)
[2024-06-26 18:15:02,753 - ERROR - fouram]: RPC error: [create_index], <MilvusException: (code=1, message=450737097756315587: failed to create index, C Runtime Exception: index not supported
: segcore unsupported error[segcoreCode=2003];)>, <Time:{'RPC start': '2024-06-26 18:15:02.243028', 'RPC error': '2024-06-26 18:15:02.753373'}> (decorators.py:146)
[2024-06-26 18:15:02,755 - ERROR - fouram]: (api_response) : [Index] <MilvusException: (code=1, message=450737097756315587: failed to create index, C Runtime Exception: index not supported
: segcore unsupported error[segcoreCode=2003];)>, [requestId: 0180e52c-33e8-11ef-b9ae-de58703b56af] (api_request.py:57)
[2024-06-26 18:15:02,755 - ERROR - fouram]: [CheckFunc] init_index request check failed, response:<MilvusException: (code=1, message=450737097756315587: failed to create index, C Runtime Exception: index not supported
: segcore unsupported error[segcoreCode=2003];)> (func_check.py:54)

milvus log:
image

Expected Behavior

No response

Steps To Reproduce

No response

Milvus Log

No response

Anything else?

No response

@elstic elstic added kind/bug Issues or changes related a bug needs-triage Indicates an issue or PR lacks a `triage/foo` label and requires one. test/benchmark benchmark test labels Jun 27, 2024
@xiaocai2333
Copy link
Contributor

@elstic
Did you enable diskann when building the image?

@elstic
Copy link
Contributor Author

elstic commented Jun 27, 2024

@elstic Did you enable diskann when building the image?

The image is automatically typed by pipeline, I looked for the log that typed the image.

image

@yanliang567
Copy link
Contributor

/assign @xiaocai2333
/unassign

@yanliang567 yanliang567 added triage/accepted Indicates an issue or PR is ready to be actively worked on. and removed needs-triage Indicates an issue or PR lacks a `triage/foo` label and requires one. labels Jun 27, 2024
@yanliang567 yanliang567 added this to the 2.4.6 milestone Jun 27, 2024
@wangting0128
Copy link
Contributor

wangting0128 commented Jun 27, 2024

same error

argo task: multi-vector-corn-1-1719410400
test case name: test_hybrid_search_locust_dml_load_release_cluster
image:2.4-20240626-07a05db5-amd64

server:

NAME                                                              READY   STATUS                            RESTARTS        AGE     IP              NODE         NOMINATED NODE   READINESS GATES
multi-vector-corn-1-1719410400-3-etcd-0                           1/1     Running                           0               7m25s   10.104.34.252   4am-node37   <none>           <none>
multi-vector-corn-1-1719410400-3-etcd-1                           1/1     Running                           0               7m24s   10.104.25.74    4am-node30   <none>           <none>
multi-vector-corn-1-1719410400-3-etcd-2                           1/1     Running                           0               7m24s   10.104.20.117   4am-node22   <none>           <none>
multi-vector-corn-1-1719410400-3-milvus-datacoord-674f9f45jtsqg   1/1     Running                           4 (5m43s ago)   7m25s   10.104.13.115   4am-node16   <none>           <none>
multi-vector-corn-1-1719410400-3-milvus-datanode-856995b7dhmcsz   1/1     Running                           4 (5m49s ago)   7m25s   10.104.24.101   4am-node29   <none>           <none>
multi-vector-corn-1-1719410400-3-milvus-datanode-856995b7dqvcfh   1/1     Running                           4               7m25s   10.104.25.58    4am-node30   <none>           <none>
multi-vector-corn-1-1719410400-3-milvus-indexcoord-564fddbngtvf   1/1     Running                           0               7m25s   10.104.13.114   4am-node16   <none>           <none>
multi-vector-corn-1-1719410400-3-milvus-indexnode-555c49d4jj7kr   1/1     Running                           4 (6m1s ago)    7m25s   10.104.14.77    4am-node18   <none>           <none>
multi-vector-corn-1-1719410400-3-milvus-proxy-6569f58fc4-4l6ff    1/1     Running                           4 (5m59s ago)   7m25s   10.104.6.106    4am-node13   <none>           <none>
multi-vector-corn-1-1719410400-3-milvus-querycoord-79f9f74mltjt   1/1     Running                           4 (5m59s ago)   7m25s   10.104.6.109    4am-node13   <none>           <none>
multi-vector-corn-1-1719410400-3-milvus-querynode-55db44b6bk2wk   1/1     Running                           4 (5m45s ago)   7m25s   10.104.34.236   4am-node37   <none>           <none>
multi-vector-corn-1-1719410400-3-milvus-querynode-55db44b6sh4mj   1/1     Running                           4 (5m49s ago)   7m25s   10.104.18.224   4am-node25   <none>           <none>
multi-vector-corn-1-1719410400-3-milvus-rootcoord-79749876pn2gr   1/1     Running                           4 (6m ago)      7m25s   10.104.14.78    4am-node18   <none>           <none>
multi-vector-corn-1-1719410400-3-minio-0                          1/1     Running                           0               7m25s   10.104.34.251   4am-node37   <none>           <none>
multi-vector-corn-1-1719410400-3-minio-1                          1/1     Running                           0               7m24s   10.104.20.113   4am-node22   <none>           <none>
multi-vector-corn-1-1719410400-3-minio-2                          1/1     Running                           0               7m24s   10.104.25.75    4am-node30   <none>           <none>
multi-vector-corn-1-1719410400-3-minio-3                          1/1     Running                           0               7m23s   10.104.24.115   4am-node29   <none>           <none>
multi-vector-corn-1-1719410400-3-pulsar-bookie-0                  1/1     Running                           0               7m24s   10.104.16.95    4am-node21   <none>           <none>
multi-vector-corn-1-1719410400-3-pulsar-bookie-1                  1/1     Running                           0               7m23s   10.104.25.79    4am-node30   <none>           <none>
multi-vector-corn-1-1719410400-3-pulsar-bookie-2                  1/1     Running                           0               7m23s   10.104.20.118   4am-node22   <none>           <none>
multi-vector-corn-1-1719410400-3-pulsar-bookie-init-r2ggt         0/1     Completed                         0               7m25s   10.104.5.246    4am-node12   <none>           <none>
multi-vector-corn-1-1719410400-3-pulsar-broker-0                  1/1     Running                           0               7m25s   10.104.20.104   4am-node22   <none>           <none>
multi-vector-corn-1-1719410400-3-pulsar-proxy-0                   1/1     Running                           0               7m24s   10.104.5.247    4am-node12   <none>           <none>
multi-vector-corn-1-1719410400-3-pulsar-pulsar-init-2cbwc         0/1     Completed                         0               7m25s   10.104.6.107    4am-node13   <none>           <none>
multi-vector-corn-1-1719410400-3-pulsar-recovery-0                1/1     Running                           0               7m25s   10.104.6.108    4am-node13   <none>           <none>
multi-vector-corn-1-1719410400-3-pulsar-zookeeper-0               1/1     Running                           0               7m25s   10.104.34.250   4am-node37   <none>           <none>
multi-vector-corn-1-1719410400-3-pulsar-zookeeper-1               1/1     Running                           0               5m55s   10.104.20.145   4am-node22   <none>           <none>
multi-vector-corn-1-1719410400-3-pulsar-zookeeper-2               1/1     Running                           0               4m20s   10.104.17.28    4am-node23   <none>           <none> 

client pod name: multi-vector-corn-1-1719410400-3362665307
client log:
image

test steps:

        concurrent test and calculation of RT and QPS

        :purpose:  `DML & load -> release collection`
            verify DML & load -> release collection scenario,
            which has 4 vector fields(IVF_FLAT, HNSW, DISKANN, IVF_SQ8) and scalar fields: `int64_1`, `varchar_1`

        :test steps:
            1. create collection with fields:
                'float_vector': 128dim,
                'float_vector_1': 128dim,
                'float_vector_2': 128dim,
                'float_vector_3': 128dim,
                scalar field: int64_1, varchar_1
            2. build indexes:
                IVF_FLAT: 'float_vector'
                HNSW: 'float_vector_1',
                DISKANN: 'float_vector_2'
                IVF_SQ8: 'float_vector_3'
                INVERTED: 'int64_1', 'varchar_1'
                default scalar index: 'id'
            3. insert 100k data
            4. flush collection
            5. build indexes again using the same params <- build diskann failed

@xiaofan-luan
Copy link
Contributor

on build.sh

UBUNTU system build diskann index

if [ "$OS_NAME" == "ubuntu20.04" ] ; then
BUILD_DISK_ANN=ON
fi

since we upgrade to 22.04, that might be issue

@wangting0128
Copy link
Contributor

wangting0128 commented Jun 27, 2024

on build.sh

UBUNTU system build diskann index

if [ "$OS_NAME" == "ubuntu20.04" ] ; then BUILD_DISK_ANN=ON fi

since we upgrade to 22.04, that might be issue

@samhuang-z

please help check it out~

@xiaofan-luan
Copy link
Contributor

OS_NAME=$(uname -s | tr '[:upper:]' '[:lower:]')

OS_VERSION="unknown"
DISTRO_NAME="unknown"

if [[ -f /etc/os-release ]]; then
. /etc/os-release
DISTRO_NAME=$ID
OS_VERSION=$VERSION_ID
fi

if [[ ("$DISTRO_NAME" == "ubuntu" && "$OS_VERSION" >= "20.04") || "$DISTRO_NAME" == "rocky" || "$DISTRO_NAME" == "amzn" ]]; then
BUILD_DISK_ANN=ON
fi

@samhuang-z can you help me to verify the code above work?

@samhuang-z
Copy link
Collaborator

@xiaofan-luan It's quite crazy and challenging to handle float and complex if statement in shell. how about this? make it more easy

sre-ci-robot pushed a commit that referenced this issue Jun 27, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
kind/bug Issues or changes related a bug test/benchmark benchmark test triage/accepted Indicates an issue or PR is ready to be actively worked on.
Projects
None yet
Development

No branches or pull requests

6 participants