#2626 [NPU训练任务] 发现用户可以同时启动2个NPU训练任务

Closed
created 1 year ago by deng · 7 comments
deng commented 1 year ago
发现有平台用户可以同时启动2个NPU训练任务。 任务名:Rober202207291164934 ; Rober202207291164934
deng added the
question
label 1 year ago
zhoupzh was assigned by liuzx 1 year ago
tanglj added this to the V20220830 milestone 1 year ago
liuzx was assigned by tanglj 1 year ago
tanglj commented 1 year ago
Collaborator
与 #2069 aiforge创建调试任务,等待非常久,而且出现两个一模一样的任务 重复 https://git.openi.org.cn/OpenI/aiforge/issues/2069
zhoupzh added the
test
label 1 year ago
wangj commented 1 year ago
Owner
未从根源上解决,仍然能复现。 taskA在界面提交后未入库前,再次提交创建taskA 是可以提交成功的。 ![image](/attachments/7dcfc48e-df00-4d57-a1ed-425d1697c7b7)
wangj removed the
test
label 1 year ago
liuzx commented 1 year ago
Collaborator
在fix-2069单独拉了分支,后台采用保存rediskey的方案,可以测试一下
liuzx added the
test
label 1 year ago
wangj commented 1 year ago
Owner
移到下个里程碑
wangj modified the milestone from V20220830 to V20220908 1 year ago
wangj commented 1 year ago
Owner
目前的效果是这样。 同1个账号: 1.同类型云脑任务,同一个项目下可以同时启动2个相同名称的 [fixed。有待测试] 2.同类型云脑任务,同一个项目下可以同时启动2个不同名称的 [不涉及。仍然可以] 3.同类型云脑任务,不同项目下可以同时启动2个,名称相同 [不涉及。仍然可以] 4.同类型云脑任务,不同项目下可以同时启动2个,名称不同 [不涉及。仍然可以] 确认一下这样修改是否符合需求?是否需要评审技术方案?@tanglj 针对#1, 已发现问题: #2858 , #2859
wangj added the
need review
label 1 year ago
tanglj was assigned by wangj 1 year ago
wangj modified the milestone from V20220908 to V20220926 1 year ago
wangj commented 1 year ago
Owner
> 目前的效果是这样。 > 同1个账号: > 1.同类型云脑任务,同一个项目下可以同时启动2个相同名称的 [fixed。有待测试] > 2.同类型云脑任务,同一个项目下可以同时启动2个不同名称的 [不涉及。仍然可以] > 3.同类型云脑任务,不同项目下可以同时启动2个,名称相同 [不涉及。仍然可以] > 4.同类型云脑任务,不同项目下可以同时启动2个,名称不同 [不涉及。仍然可以] > 确认一下这样修改是否符合需求?是否需要评审技术方案?@tanglj > > 针对#1, 已发现问题: #2858 , #2859 已经评审,只处理情况1。由于云脑2已经下线,暂时无法验证npu训练场景。
wangj removed the
need review
label 1 year ago
wangj commented 1 year ago
Owner
移到下个里程碑,等云脑2上线了 验证。
wangj modified the milestone from V20220926 to V20221019 1 year ago
liwei03 closed this issue 1 year ago
Sign in to join this conversation.
No Milestone
No Assignees
4 Participants
Notifications
Due Date

No due date set.

Dependencies

This issue currently doesn't have any dependencies.

Loading…
There is no content yet.