leofs_test2
is a integration test tool for LeoFS clusters. Tests/Scenarios depend on Erlang and can communicate with the cluster using the distributed Erlang.
- "leofs_test2" requires Erlang R16B03-1 or later.
- "leofs_test2" uses the rebar build system. Makefile so that simply running "make" at the top level should work.
$ git clone https://github.com/leo-project/leofs_test2.git
$ cd leofs_test2
$ make
- Add a domain for the LeoFS bucket in
/etc/hosts
$ sudo vi /etc/hosts
## Replace {BUCKET_NAME} with the name of the bucket ##
127.0.0.1 localhost <BUCKET_NAME>.localhost
- Build a LeoFS of your test with "bootstrap.sh" which is included in LeoFS source code
$ cd <leofs-root-dir>/
$ sh bootstrap.sh build integration-test
- leofs_test2 has options as follows. You can set suitable option(s) with tests.
$ ./leofs_test --help
Usage: leofs_test [-b <bucket>] [-c <cookie>] [-d <leofs_dir>] [-k <keys>] [-m <manager>] [-t <test>] [-h] [-v]
-b, --bucket Target a bucket
-c, --cookie Distributed-cookie for communication with LeoFS
-d, --dir LeoFS directory
-k, --keys Total number of keys
-m, --manager LeoFS Manager
-t, --test Execute a test
-h, --help Show the program options
-v, --version Show version information
-
Default value of options
- bucket: "backup"
- cookie: "401321b4"
- keys: 10000
- manager: '[email protected]'
-
Example - Execute the scenarios
$ ./leofs_test -d <leofs-package-dir> -b <bucket> -k 10000 -c leofs_test
...
[{bucket,"test"},
{leofs_dir,"/home/leofs/dev/projects/leo-project/leofs/package/"}]
::: START :::
[SCENARIO-1]
* mnesia (test mnesia backup/restore)
* user_crud (test user related CRUD)
* endpoint_crud (test endpoint related CRUD)
* bucket_crud (test bucket related CRUD)
* watch_mq (watch state of mq)
#1
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* update_log_level (update log level of a node)
* dump_ring (dump ring data to the local disk)
* create_bucket (create a bucket)
* update_consistency_level (update consistency level of a node)
* purge_cache (remove a cache from each gateway)
* recover_file (recover data of a file)
#1
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
#1
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* get_objects_not_found (get objects_not_found)....................................................................................................|
* put_zero_byte_objects (put zero byte objects)....................................................................................................|
* get_objects (get objects)....................................................................................................|
* put_objects (put objects)....................................................................................................|
* get_objects (get objects)....................................................................................................|
* check_redundancies (check redundancies of replicas)....................................................................................................|
* del_objects (remove objects)....................................................................................................|
* check_redundancies (check redundancies of replicas)....................................................................................................|
* mp_upload_normal (multipart upload)
* mp_upload_normal_in_parallel (multipart upload in parallel)
* mp_upload_abort (abort multipart upload)
#1
* '[email protected]':[{leo_delete_dir_queue_1,4}]
* '[email protected]':[{leo_delete_dir_queue_1,2}]
* '[email protected]':[{leo_delete_dir_queue_1,4}]
* '[email protected]':[{leo_delete_dir_queue_1,3}]
#2
* '[email protected]':[{leo_delete_dir_queue_1,2}]
* '[email protected]':[{leo_delete_dir_queue_1,1}]
* '[email protected]':[{leo_delete_dir_queue_1,1}]
* '[email protected]':[{leo_delete_dir_queue_1,3}]
#3
* '[email protected]':[{leo_delete_dir_queue_1,2}]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
#4
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* mp_upload_invalid_complete (invalid complete multipart upload)
.
.
.
[SCENARIO-7]
* delete_bucket (delete a bucket)
* watch_mq (watch state of mq)
#1
* '[email protected]':[{leo_delete_dir_queue_1,3165}]
* '[email protected]':[{leo_delete_dir_queue_1,4399}]
* '[email protected]':[{leo_delete_dir_queue_1,2679}]
* '[email protected]':[{leo_delete_dir_queue_1,1844}]
#2
* '[email protected]':[{leo_delete_dir_queue_1,442}]
* '[email protected]':[{leo_delete_dir_queue_1,713}]
* '[email protected]':[{leo_delete_dir_queue_1,208}]
* '[email protected]':[{leo_delete_dir_queue_1,1012}]
#3
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[{leo_delete_dir_queue_1,609}]
#4
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* create_bucket (create a bucket)
* get_objects_not_found (get objects_not_found)....................................................................................................|
* put_inconsistent_objects (put inconsistent objects)....................................................................................................|
* scrub_cluster (scrub the whole cluster)
#1
* '[email protected]':[{leo_per_object_queue,4470}]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
#2
* '[email protected]':[{leo_per_object_queue,2552}]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
#3
* '[email protected]':[{leo_per_object_queue,13}]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
#4
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* '[email protected]':[N/A]
* check_redundancies (check redundancies of replicas)....................................................................................................|
::: Finished (817sec) :::
- Example - Execute only a test
$ ./leofs_test -d leofs/package -t check_redundancies
...
:: TEST: check_redundancies (check redundancies of replicas)::::
[TEST]
* check_redundancies (check redundancies of replicas)....................................................................................................|
::: Finished (13sec) :::
v0.4: Improve the basic scenarios retrieved from leofs-adm's commands Phase 1 (LeoFS 1.4.2 will come out with v0.4 tests passed)
- check whether mq-(suspend|resume|stats) works
- check whether recover-(file|node|consistency) works
- check whether compaction-(suspend|resume) works
- check whether du works
- check whether purge-cache works
- check whether (backup|restore)-mnesia works
- check whether dump-ring works
- check whether update-consistency-level works
- check whether update-log-level works
- check whether user related functions work
- check whether endpoint related functions work
- check whether bucket related functions work
- check whether rollback works
- check whether multipart uploads work
- move onto rebar3
- add the scenario-7 for the scrub cluster usecase
- check whether recover-(ring|cluster|disk)
- check whether du detail works
- check whether remove-gateway works
- check whether update-managers works
- check whether multi-dc related functions work
- check whether chown-bucket works
- check whether update-acl works
- check whether GET existing objects works while rebalance is ongoing with R=1,2
- check whether GET existing objects doesn't work while rebalance is ongoing with R=3
- check whether GET updated-after-rebalance objects works while rebalance is ongoing with R=1,2
- check whether GET updated-after-rebalance objects doesn't work while rebalance is ongoing with R=3
- check whether access_key_id and secret_access_key pairs generated by (create|import)-user works with erlcloud
- check whether GET with Range header works on both protocols(S3 and REST)
- check whether leo_storage cluster works with one storage being dead
- check whether leo_storage cluster fail-back
- check whether leo_storage cluster respond an expected error in case # of replicas < N
- check whether leo_storage cluster works while compaction is ongoing
- check whether leo_storage cluster works while rebalance is ongoing
- check whether leo_storage cluster works while compaction AND rebalance are ongoing
LeoProject/LeoFS is sponsored by Rakuten, Inc. and Rakuten Institute of Technology.