[
https://issues.apache.org/jira/browse/MESOS-970?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15394431#comment-15394431
]
Bing Li commented on MESOS-970:
-------------------------------
I see. "-j" does make compile faster. I thought it would help the bench test
cases too.
Compared with the log you posted at
https://gist.github.com/haosdent/ce2544d92c21314571f5a7c4312eb2dc#file-mesos_1-4_bench-txt-L4
.
It seems there are much more test cases to run now.
For instance, In your log, I saw 2 times "round 0" ... to "round 399". (400 * 2
= 800 in total.)
In mine, there are at least a couple of thousands of "round n allocate() ..".
And it is "... to make 1000 offers" now, vs yours "... to make 200 offers".
At the end of your log. I saw "[==========] 74 tests from 2 test cases ran.
(51732372 ms total)". (51732372 ms = 14.37 hours).
I gave up my run because it took too long. I thought there was something wrong.
I finished the run a couple of weeks ago. But now it will take much longer.
I'll rerun my bench test and share the log.
Is the super long bench test as expected?
Do you mind running the latest master branch to have a look?
My concern is that since the test cases changed quite a lot, it won't be a fair
comparison between leveldb 1.4 and 1.18.
I also have 8 core X86 machine running master branch. It is not fast either.
Thanks,
> Upgrade bundled leveldb to 1.18
> -------------------------------
>
> Key: MESOS-970
> URL: https://issues.apache.org/jira/browse/MESOS-970
> Project: Mesos
> Issue Type: Improvement
> Components: replicated log
> Reporter: Benjamin Mahler
> Assignee: Tomasz Janiszewski
>
> We currently bundle leveldb 1.4, and the latest version is leveldb 1.18.
> Upgrade to 1.18 could solve the problems when build Mesos in some non-x86
> architecture CPU.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
