search

eBay / griffin

Model driven data quality service
https://ebay.github.io/griffin/
Other
240 stars 165 forks source link

Model Stage is always initial status, never be executed #30

Open ljzhu1990 opened 7 years ago

ljzhu1990 commented 7 years ago

Hello, I followed the user guide.Firstly, I created a assert, and then,I created a model . But, the model was always inital status ,and it has never been executed. Can you help me solve this problem ? Thank you very much.

bhlx3lyx7 commented 7 years ago

Thanks for your question, we really want to help you resolve it. First, what's your environment, the docker env we provided or the env you built yourself? And then, can you tell us the detail of the asset you've created, we'll have a look at it. Thank you.

ljzhu1990 commented 7 years ago

I'm glad to receive your reply. I run griffin in my own environment following 'How deploy and run at local '. Firstly,I installed MongoDB , Hadoop , Spark, Hive and so on. I deployed griffin engine and hive in the same hadoop cluster. Next, I used your sample data and loaded data into hive table. Their hdfs path were : '/hivetable/user_info_src/' and '/hivetable/user_info_target' . Then, I created two data asserts which named user_info_src and user_info_target, and entered their hdfs paths each other. Next, I created a 'Accuracy' DQ model, selecting part columns from source and targert data assert for comparison. But it never execute it , and I can't get any result. There were not error info printed, too. So, Can you help me find out problems? Thank you very much.

bhlx3lyx7 commented 7 years ago

Well, I've reviewed our "How to deploy and run at local" part, and find that our steps are not described clearly, also with some version mistakes. I've modified that document now. Would you mind give us your email to discuss the detail issues? Or you can contact us through "Google Groups" mentioned here. Thanks a lot.

ljzhu1990 commented 7 years ago

My email is: ljzhu1990@163.com. I will be glad to discuss the issues with you and I'm also interested in griffin. Best wishes!

titt commented 7 years ago

Hello, I have almost the same issue. I created some model but any metric was created from my model... I followed the user guide. I use a hortonworks sandbox with Spark1.6.2, hadoop2.7.3, hive1.2.1, tomcat 7.0.75.0, java1.8.0_111, mongodb3.4.3. I created users_info_src and users_info_target table and _SUCCESS file in hdfs table. I created a new directory with env.sh, griffin_jobs.sh, griffin_regular_run.sh, tmp, griffin-models.jar and I runned nohup ./griffin_regular_run.sh &. I also put griffin-core/target/ROOT.war in /usr/share/tomcat/webapps. At first I had an error 500 when I tried to create a dataassets, I put griffin-core/target/ROOT dir in /usr/share/tomcat/webapps to correct that. And now I have the same issue that describe in this case. Have you solve the case ?

Thank you very much

sorabh89 commented 7 years ago

Hi, I'm facing the same issue. I'm using the docker procedure. Also please let me know if I need to create a hive table for a file that i update in hdfs, or can it use the raw file itself for the model without a hive table. And also please update the documentation. Its hard to understand it from the documentation provided. Thanks.

luzx02 commented 7 years ago

Hi Kumer,

Griffin has been moved to Apache, please check this link.

https://github.com/apache/incubator-griffin

Thx

发自我的 iPhone

在 2017年6月8日，下午5:15，Sorabh Kumar notifications@github.com<mailto:notifications@github.com> 写道：

Hi, I'm facing the same issue. I'm using the docker procedure. Also please let me know if I need to create a hive table for a file that i update in hdfs, or can it use the raw file itself for the model without a hive table. And also please update the documentation. Its hard to understand it from the documentation provided. Thanks.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/eBay/griffin/issues/30#issuecomment-307047095, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AOZliBtuYgihWZToWPPt2zSTirdIYasJks5sB7vHgaJpZM4KoE_o.

luzx02 commented 7 years ago

Hi Kumer,

Please follow docker in incubator-griffin.

Any issue, please let us know or send email to dev@griffin.incubator.apache.orgmailto:dev@griffin.incubator.apache.org

Thanks, William

发件人: Alex Lvmailto:luzx02@hotmail.com 发送时间: 2017年6月8日 18:30 收件人: eBay/griffinmailto:reply@reply.github.com 抄送: eBay/griffinmailto:griffin@noreply.github.com; Subscribedmailto:subscribed@noreply.github.com; dev@griffin.incubator.apache.orgmailto:dev@griffin.incubator.apache.org 主题: Re: [eBay/griffin] Model Stage is always initial status, never be executed (#30)

Hi Kumer,

Griffin has been moved to Apache, please check this link.

https://github.com/apache/incubator-griffin

Thx

发自我的 iPhone

在 2017年6月8日，下午5:15，Sorabh Kumar notifications@github.com<mailto:notifications@github.com> 写道：

Hi, I'm facing the same issue. I'm using the docker procedure. Also please let me know if I need to create a hive table for a file that i update in hdfs, or can it use the raw file itself for the model without a hive table. And also please update the documentation. Its hard to understand it from the documentation provided. Thanks.

― You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/eBay/griffin/issues/30#issuecomment-307047095, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AOZliBtuYgihWZToWPPt2zSTirdIYasJks5sB7vHgaJpZM4KoE_o.