incubator-uniffle
incubator-uniffle copied to clipboard
Execute start script with nohup
What changes were proposed in this pull request?
Execute start script with nohup
Why are the changes needed?
Process don't exit if exec start script using ansible. Therefore, we can't do batch start operation for this.
Does this PR introduce any user-facing change?
No
How was this patch tested?
Have already verify in our production enviroment.
Codecov Report
Merging #110 (698875e) into master (deb7fe4) will not change coverage. The diff coverage is
n/a
.
@@ Coverage Diff @@
## master #110 +/- ##
=========================================
Coverage 56.52% 56.52%
Complexity 1183 1183
=========================================
Files 149 149
Lines 8019 8019
Branches 767 767
=========================================
Hits 4533 4533
Misses 3242 3242
Partials 244 244
Help us with your feedback. Take ten seconds to tell us how you rate us.
@zuston I don't have ansible environment, Could you help me review and verify this path?
I use the shell module of ansible to start the shuffle server without this PR, it wont block. command as: ansible -i hosts_xxx -m shell -a "./start-shuffle-server.sh"
Could u help provide some detailed info that which ansible module dont work? I will test it. @xianjingfeng
I use the shell module of ansible to start the shuffle server without this PR, it wont block. command as:
ansible -i hosts_xxx -m shell -a "./start-shuffle-server.sh"
Could u help provide some detailed info that which ansible module dont work? I will test it. @xianjingfeng
I forget. I'll ask my colleague next week. I tried it myself just now, and i found the jvm will exit immediately without this pr.
@zuston Blocked is because we have made a layer of encapsulation for ansible. We will check if the process start successful. Process will not start successful by using native ansible without this pr.
But the process could be started successfully by native ansible for me.
If this pr only solve the issue of company's internal specific implement, I prefer not merging it.
@jerqi @zuston
@zuston What's your suggestion?
Hadoop, Spark, Hive, and Trino support run service both on frontend or as daemon, can we follow them?
I also use the ansible of 2.9.27 but no such problems. @xianjingfeng And i dont see any blocking in your sceenshot.
Hadoop, Spark, Hive, and Trino support run service both on frontend or as daemon, can we follow them?
Agree with u @pan3793
And i dont see any blocking in your sceenshot.
Not blocked, jvm process have not start if start by ansible
Got your point. @xianjingfeng But i think we should dig the cause rather than using the nohup directly? But i have no solutions currently