fluo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From GitBox <...@apache.org>
Subject [GitHub] kpm1985 commented on issue #1004: FLUO-1000 OracleServer race conditions
Date Thu, 01 Jan 1970 00:00:00 GMT
kpm1985 commented on issue #1004: FLUO-1000 OracleServer race conditions
URL: https://github.com/apache/fluo/pull/1004#issuecomment-359928917
 
 
   Apologies for the quality of type to text doesn't seem to be showing me
   much love today.
   
   I saw the curator issue that was put out and took a look at it that's cool.
   Wanted to say that I did throw a few statements in to print the curator
   statement before exceptions were thrown. I saw during all of the exceptions
   the state of the curator framework was stopped not latent but stopped.
   
   I think most of these errors are happening due to the 5th test in OracleIT.
   It does a lot of oserver.stop oserver.start stuff.
   
   Apologies for the type to text.
   
   On Jan 23, 2018 12:56 PM, "Kenneth Mcfarland" <kennethmcfarland@apache.org>
   wrote:
   
   > Hi Keith,
   >
   > About the errors that show during my verify. My original solution I sit
   > and check curator state inside of the function takeLadership and wait until
   > its ready.
   >
   > I tried moving this check to the start method and I ran Maven verify on
   > this a few times and to my surprise I kept getting some exceptions on the
   > state of that curator framework again.
   >
   > So what I did is I left it in that method that is called by leader
   > selector and when the check is in that method with a couple other checks
   > like this synchronized on the leader variable then the errors are not
   > present.
   >
   > This made me wonder after tracing the flow of what I thought was going on
   > in Oracle server, well what the heck is happening somebody else is screwing
   > with something in the background thread somewhere so, I still didn't really
   > understand what was going on but I left the check inside of the i, I still
   > didn't really understand what was going on but I left the check inside of
   > the is take leadership method.
   >
   > My only thought is possibly multiple instances but I thought those were
   > private variables so not exactly sure what's going on with that.
   >
   > Anyways I really appreciate you responding and looking into this with me
   > this is the help I totally needed. I'm on break with the phone so I
   > apologize for not highlighting code and doing other things that you can do
   > on a computer.
   >
   > On Jan 23, 2018 8:42 AM, "Keith Turner" <notifications@github.com> wrote:
   >
   >> Maybe we don't need to verify that the illegal state came from curator.
   >> We could make the check stronger, something like the following.
   >>
   >>  } catch (IllegalStateException e) {
   >>       if(curatorFramework.getState() == STOPPED && !started) {
   >>         // the oracle server had been stopped, do not care much about errors...
lets suppress the exception
   >>         log.debug(...);  log a debug message that this happened
   >>       } else {
   >>         throw e;
   >>       }
   >>     }
   >>
   >> ?
   >> You are receiving this because you were mentioned.
   >> Reply to this email directly, view it on GitHub
   >> <https://github.com/apache/fluo/pull/1004#issuecomment-359852270>, or mute
   >> the thread
   >> <https://github.com/notifications/unsubscribe-auth/Acg-LI5CI5WUf9Avp_CQOuOa08pi5s8Cks5tNgv9gaJpZM4Rh_58>
   >> .
   >>
   >
   

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services

Mime
View raw message