[LU-6225] test-framework does not cleanup the env for failed tests

• Type: Bug
• Resolution: Fixed
• Priority: Major
• Fix Version/s: Lustre 2.8.0
• Affects Version/s: Lustre 2.5.3, Lustre 2.8.0

Epic/Theme

patch test

• Severity: 3
• Project: Test Infrastructure
• Assignee: WC Triage
• Reporter: gaurav mahajan (Inactive)

Description

if a test fails with an error, we just exit:

error() {
	error_noexit "$@"
	exit 1
}

and therefore the end of run_one() does not execute, what means we do not cleanup the env settings (e.g. fail_loc) before running the next test.

Issue Links

• is duplicated by: LU-2312 fail_loc is not reset if error() is called

Comments

[Joseph Gmitter (Inactive)]

Landed for 2.8

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch http://review.whamcloud.com/13692/
Subject: LU-6225 test: test-framework does not cleanup for failed tests
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: dc3a6a13d2cec51ca5d8ab1c870ce4cee3d80494

[Saurabh Tandan (Inactive)]

Encountered same issue for tag 2.7.66 for FULL- EL7.1 Server/EL6.7 Client , master , build# 3314
https://testing.hpdd.intel.com/test_sets/924b1b78-ca91-11e5-9609-5254006e85c2

Another failure for master : Tag 2.7.66 FULL - EL7.1 Server/SLES11 SP3 Client, build# 3314
https://testing.hpdd.intel.com/test_sets/b0f05d56-ca7b-11e5-9609-5254006e85c2

[Saurabh Tandan (Inactive)]

Another Instance for interop: EL7 Server/2.7.1 Client
Server: master, build# 3303, RHEL 7
Client: 2.7.1, b2_7_fe/34
https://testing.hpdd.intel.com/test_sets/7a715f8a-bac4-11e5-9137-5254006e85c2
https://testing.hpdd.intel.com/test_sets/7a2d8fe4-bac4-11e5-9137-5254006e85c2

[Saurabh Tandan (Inactive)]

Another instance found for :
Server: Master , Build# 3276
Client: b2_7_fe/34
https://testing.hpdd.intel.com/test_sets/1c4ee028-a5a5-11e5-a14a-5254006e85c2
https://testing.hpdd.intel.com/test_sets/1c77ef4a-a5a5-11e5-a14a-5254006e85c2

[Saurabh Tandan (Inactive)]

Server: Master, Build# 3266, Tag 2.7.64 , RHEL 7
Client: 2.5.5, b2_5_fe/62
https://testing.hpdd.intel.com/test_sets/0d665f82-a000-11e5-a33d-5254006e85c2
https://testing.hpdd.intel.com/test_sets/0d289292-a000-11e5-a33d-5254006e85c2

[Sarah Liu]

Hit in interop testing
https://testing.hpdd.intel.com/test_sessions/2b6be11e-9033-11e5-a833-5254006e85c2
server: lustre-master/3251 EL7
client 2.7.0

[gaurav mahajan (Inactive)]

The tests which call error_noexit() will return with a value (non zero value) to run_one(). Then run_one() will call error() in case of non zero return value from the test which will again call error_noexit().

If we include reset_fail_loc to error_noexit () then it becomes kind of doing the things which are already done as error() will ultimately get called incase of non zero return value (ie on error)

In case of error_and_remount(), agree that it calls error_no_exit() and exits from there. I guess modification is needed for test framework to handle this looping of error() and error_no_exit().

As of now, I will accept the comments and move reset_fail_loc to error_no_exit(). But modifications are needed for test_framework.sh

Please leave your comments on this.

[Gerrit Updater]

Andreas Dilger ([e-mail]) uploaded a new patch: http://review.whamcloud.com/13692
Subject: LU-6225 test: test-framework does not cleanup for failed tests
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: 00afee36ddfb1fb7bb4b58ca82b0fef41f7f4527

[Peter Jones]

Does this issue only affect b2_5 or is it relevant for master too?