Xbeech
Level 1 (0 points)

Q: doveadm process going nutso

I'm running Server 5.015, and the computer on which it is running got to be very sluggish a couple of software updates ago.  I looked in Console and found the following in the logs - happening many times every second, and Activity Monitor shows Python and doveadm regularly using 80%+ CPU:

 

2/17/16 4:59:23.000 PM kernel[0]: Sandbox: doveadm(83371) deny(1) mach-lookup com.apple.lsd.mapdb

2/17/16 4:59:23.055 PM doveadm[83371]: LaunchServices: received XPC_ERROR_CONNECTION_INVALID trying to map database

2/17/16 4:59:23.055 PM doveadm[83371]: LaunchServices: disconnect event received for service com.apple.lsd.mapdb

2/17/16 4:59:23.131 PM servermgr_dns[83618]: validating connection from 0 : 100000

2/17/16 4:59:23.000 PM kernel[0]: Sandbox: doveadm(83371) deny(1) mach-lookup com.apple.lsd.mapdb

2/17/16 4:59:23.000 PM kernel[0]: Sandbox: doveadm(83371) deny(1) mach-lookup com.apple.lsd.mapdb

2/17/16 4:59:23.000 PM kernel[0]: Sandbox: doveadm(83371) deny(1) mach-lookup com.apple.lsd.mapdb

2/17/16 4:59:23.000 PM kernel[0]: Sandbox: doveadm(83371) deny(1) mach-lookup com.apple.lsd.mapdb

2/17/16 4:59:23.000 PM kernel[0]: Sandbox: doveadm(83371) deny(1) mach-lookup com.apple.lsd.mapdb

2/17/16 4:59:23.000 PM kernel[0]: Sandbox: doveadm(83371) deny(1) mach-lookup com.apple.lsd.mapdb

2/17/16 4:59:23.000 PM kernel[0]: Sandbox: doveadm(83371) deny(1) mach-lookup com.apple.lsd.mapdb

2/17/16 4:59:23.000 PM kernel[0]: Sandbox: doveadm(83371) deny(1) mach-lookup com.apple.lsd.mapdb

2/17/16 4:59:23.000 PM kernel[0]: Sandbox: doveadm(83371) deny(1) mach-lookup com.apple.lsd.mapdb

 

Any thoughts are MUCH appreciated!

Posted on Feb 17, 2016 5:05 PM

by Xbeech,Solvedanswer
Xbeech Xbeech
Level 1 (0 points)
A:

For anyone who happens upon this thread, Linc was right, of course.  I had hoped the fault was in the hard drive, which I replaced.  Yesterday, I bought a mini, did a nuke and pave, manually moving the server data, and poof: problem gone. 

Posted on Jun 2, 2016 9:18 AM

Close
Xbeech

Q: doveadm process going nutso

  • All replies
  • Helpful answers

Previous Page 2

  • by Xbeech,

    Xbeech Xbeech Mar 4, 2016 11:10 AM in response to Linc Davis
    Level 1 (0 points)
    Mar 4, 2016 11:10 AM in response to Linc Davis

    Thanks - I really meant to ask if there's a way to figure out which component of the hardware is likely the culprit.  I'll start with the startup drive and move on from there.  My plan is to install a fresh copy of OSX, use migration assistant to move the data from the existing drive, then install a fresh copy of Server.  If you know a better path, or pitfalls to avoid, I'd love to hear about them.   Thanks again.

  • by Xbeech,

    Xbeech Xbeech Mar 8, 2016 2:15 PM in response to Xbeech
    Level 1 (0 points)
    Mar 8, 2016 2:15 PM in response to Xbeech

    I got a external USB drive, installed a fresh copy of El Capitan on it.  When I started the Xserve, it was (qualitatively) very responsive.  Then I used migration assistant to transfer the server data from the startup drive. When that was complete and the server restarted, it was sluggish again, and the error logs showed all the same problems.  You may be right that it's a hardware problem, if it is, it's not the drive...

  • by Xbeech,

    Xbeech Xbeech May 31, 2016 4:58 AM in response to Linc Davis
    Level 1 (0 points)
    May 31, 2016 4:58 AM in response to Linc Davis

    Hi Linc,

      I plugged in an external drive and installed El Capitan.  Then I used migration assistant to migrate the server data.  It all started up ok, but I shortly started getting the same dovecot errors.  More importantly, I have another problem now - whenever I turn on the mail server, the named process starts connecting to hundreds and hundreds of other servers on port 53.  My firewall starts balking at 1000+ sessions.  Am I a bot?  And advice?  Thanks again.

  • by Xbeech,Solvedanswer

    Xbeech Xbeech Jun 2, 2016 9:18 AM in response to Xbeech
    Level 1 (0 points)
    Jun 2, 2016 9:18 AM in response to Xbeech

    For anyone who happens upon this thread, Linc was right, of course.  I had hoped the fault was in the hard drive, which I replaced.  Yesterday, I bought a mini, did a nuke and pave, manually moving the server data, and poof: problem gone. 

Previous Page 2