Hi Gopal,
> I have upgrade hive version to 3.0 and the somaxconn value of shuffle 
> port(15551) has been 16384 not 50. Thank you very much.
> But I encounter the following problem when run llap, and this is same with 
> https://issues.apache.org/jira/browse/HIVE-10693 . Whether it is a bug of 
> llap in hive 3.0?
2017-11-28T22:25:41,875 WARN  [Fetcher_B {Map_9} #0 ()] 
org.apache.tez.runtime.library.common.shuffle.Fetcher: Fetch Failure from host 
while connecting: bdpe91, attempt: InputAttemptIdentifier [inputIdentifier=0, 
attemptNumber=0, pathComponent=attempt_1511868736816_0074_2_00_000000_0_10003, 
spillType=0, spillId=-1] Informing ShuffleManager:
java.net.SocketTimeoutException: Read timed out
        at java.net.SocketInputStream.socketRead0(Native Method) ~[?:1.8.0_112]
        at java.net.SocketInputStream.socketRead(SocketInputStream.java:116) 
~[?:1.8.0_112]
        at java.net.SocketInputStream.read(SocketInputStream.java:170) 
~[?:1.8.0_112]
        at java.net.SocketInputStream.read(SocketInputStream.java:141) 
~[?:1.8.0_112]
        at java.io.BufferedInputStream.fill(BufferedInputStream.java:246) 
~[?:1.8.0_112]
        at java.io.BufferedInputStream.read1(BufferedInputStream.java:286) 
~[?:1.8.0_112]
        at java.io.BufferedInputStream.read(BufferedInputStream.java:345) 
~[?:1.8.0_112]
        at sun.net.www.http.HttpClient.parseHTTPHeader(HttpClient.java:704) 
~[?:1.8.0_112]
        at sun.net.www.http.HttpClient.parseHTTP(HttpClient.java:647) 
~[?:1.8.0_112]
        at 
sun.net.www.protocol.http.HttpURLConnection.getInputStream0(HttpURLConnection.java:1569)
 ~[?:1.8.0_112]
        at 
sun.net.www.protocol.http.HttpURLConnection.getInputStream(HttpURLConnection.java:1474)
 ~[?:1.8.0_112]
        at 
org.apache.tez.http.HttpConnection.getInputStream(HttpConnection.java:260) 
~[tez-runtime-library-0.9.1-SNAPSHOT.jar:0.9.1-SNAPSHOT]
        at 
org.apache.tez.runtime.library.common.shuffle.Fetcher.setupConnection(Fetcher.java:529)
 ~[tez-runtime-library-0.9.1-SNAPSHOT.jar:0.9.1-SNAPSHOT]
        at 
org.apache.tez.runtime.library.common.shuffle.Fetcher.doHttpFetch(Fetcher.java:561)
 ~[tez-runtime-library-0.9.1-SNAPSHOT.jar:0.9.1-SNAPSHOT]
at 
org.apache.tez.runtime.library.common.shuffle.Fetcher.doHttpFetch(Fetcher.java:486)
 ~[tez-runtime-library-0.9.1-SNAPSHOT.jar:0.9.1-SNAPSHOT]
        at 
org.apache.tez.runtime.library.common.shuffle.Fetcher.callInternal(Fetcher.java:284)
 ~[tez-runtime-library-0.9.1-SNAPSHOT.jar:0.9.1-SNAPSHOT]
        at 
org.apache.tez.runtime.library.common.shuffle.Fetcher.callInternal(Fetcher.java:76)
 ~[tez-runtime-library-0.9.1-SNAPSHOT.jar:0.9.1-SNAPSHOT]
        at org.apache.tez.common.CallableWithNdc.call(CallableWithNdc.java:36) 
~[tez-common-0.9.1-SNAPSHOT.jar:0.9.1-SNAPSHOT]
        at java.util.concurrent.FutureTask.run(FutureTask.java:266) 
~[?:1.8.0_112]
        at 
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142) 
[?:1.8.0_112]
        at 
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617) 
[?:1.8.0_112]
        at java.lang.Thread.run(Thread.java:745) [?:1.8.0_112]

-----Original Message-----
From: Gopal Vijayaraghavan [mailto:[email protected]] 
Sent: Tuesday, November 28, 2017 12:32 AM
To: [email protected]
Subject: Re: Hive +Tez+LLAP does not have obvious performance improvement than 
HIVE + Tez

Hi,

If you've got the 1st starvation fixed (with Hadoop 2.8 patch), all these 
configs + enable log4j2 async logging, you should definitely see a performance 
improvement.

Here's the log patches, which need a corresponding LLAP config (& have to be 
disabled in HS2, for the progress bar to work)

https://issues.apache.org/jira/browse/HIVE-13027
+
https://issues.apache.org/jira/browse/HIVE-16061

You might want to share the explain plans + a jstack, if you want me to narrow 
down on the issue.

Pick a simple query like Query55 or 52, to get a sane baseline.

> And For somaxconn, why the somaxconn value of the shuffle port (15551) is 50 
> not 16384? Thanks for your help.

Good catch, I think you might have more backports to match my GA build (which 
spikes ~80% cpu on 16 cores easily)

https://issues.apache.org/jira/browse/HIVE-16737

Cheers,
Gopal


Reply via email to