Logsi

Q: File Sharing Grinds to a Halt in Yosemite

I've been having a problem since 10.10 whereby out of the blue our Mac Mini fileserver just grinds to a halt and whenever someone tries to access files in the finder on their client Macs it will just hang and beachball.

 

I've troubleshooted a little bit from having to reboot the server to now ejecting the culprit client off the network This then speeds the network back up again but there is no way to tell which client will be the one that's caused the issue so i have to quickly run around the office and check everyones finder and the problem Mac will be the one that's a tad slower to load files in the finder than others. It's almost as if the computers are DoSing our own server which then slows down the network. There is no way to replicate this issue it seems and comes totally out of the blue.

 

We mainly have word and excel files on the server. We are connected via SMB with AFP turned off. I don't think this issue happens if we're all on AFP but different issues happen with AFP like not being able to save documents and we also need SMB on.

 

All clients are on 10.10.3 and the server is also on 10.10.3.

Posted on Apr 30, 2015 2:28 AM

Close

Q: File Sharing Grinds to a Halt in Yosemite

  • All replies
  • Helpful answers

Previous Page 2 of 4 last Next
  • by Peter Goldman1,

    Peter Goldman1 Peter Goldman1 May 29, 2015 4:46 PM in response to Logsi
    Level 1 (24 points)
    May 29, 2015 4:46 PM in response to Logsi

    Chiming in here to report that we too are seeing identical symptoms like this on *multiple* servers and separate networks.

     

    Server and clients are both running 10.10.3.

     

    Gradually SMB connections grind way way waaaaaaay down in speed.

     

    CPU and RAM utilization look great - very low.

     

    Reboot of the server "cures" it for several hours to a few days... until it occurs again, and the only thing that fixes it is another reboot.

  • by AJ 2010,

    AJ 2010 AJ 2010 May 30, 2015 9:00 AM in response to Peter Goldman1
    Level 1 (4 points)
    May 30, 2015 9:00 AM in response to Peter Goldman1

    How many clients are using SMB? Unfortunately it was such a problem for us that I had to revert back to an older 10.6 server.

  • by Peter Goldman1,

    Peter Goldman1 Peter Goldman1 Jun 2, 2015 1:25 PM in response to Logsi
    Level 1 (24 points)
    Jun 2, 2015 1:25 PM in response to Logsi

    Roughly 30 users.

     

    Not lightweight machines here too... one is a (new) Mac Pro with 128GB of ram and a 80TB fibre channel direct attached file store.   

  • by cdhw,

    cdhw cdhw Jun 2, 2015 3:00 PM in response to Peter Goldman1
    Level 4 (2,623 points)
    Servers Enterprise
    Jun 2, 2015 3:00 PM in response to Peter Goldman1

    I urge everybody who is experiencing this issue to report it to Apple as a bug:

     

         https://bugreport.apple.com/

     

    and / or open an Enterprise Support ticket. SMB is now the default file sharing protocol and it really needs to be robust enough to survive a few dozen clients opening the occasional spreadsheet...


    I suppose there's always nfsd(8). 

     

    C.

  • by AJ 2010,

    AJ 2010 AJ 2010 Jun 9, 2015 11:15 AM in response to cdhw
    Level 1 (4 points)
    Jun 9, 2015 11:15 AM in response to cdhw

    What did you write in your bug report? I'd be happy to fill this out but since it was causing so many problems I stopped using Yosemite server.

  • by Peter Goldman1,

    Peter Goldman1 Peter Goldman1 Jun 9, 2015 11:39 AM in response to AJ 2010
    Level 1 (24 points)
    Jun 9, 2015 11:39 AM in response to AJ 2010

    We were unable to pursue the issue any longer - as these were live production servers and it was causing massive issues.

     

    We reverted all users back to connecting via AFP and it is working perfectly once again.

     

    We obviously (really) need this to work in SMB in the future, but are unfortunately not in a position to troubleshoot further.

  • by AJ 2010,

    AJ 2010 AJ 2010 Jun 10, 2015 10:34 AM in response to AJ 2010
    Level 1 (4 points)
    Jun 10, 2015 10:34 AM in response to AJ 2010

    I searched the Mac Enterprise mailing list and found a few more people with this issue. Unfortunately, none of them has been able to troubleshoot it or contact Apple to get some kind of official response. I may try and reproduce this issue but I have the feeling that it has something to do with the number of clients or simultaneous connections which I won't be able to replicate.

     

    https://groups.google.com/d/msg/macenterprise/crNkLReESjE/-TXo-8Pi6C0J

     

    https://groups.google.com/d/msg/macenterprise/3XRMYNPRoWE/yGC1Gy255EwJ

  • by AJ 2010,

    AJ 2010 AJ 2010 Jun 18, 2015 4:38 PM in response to AJ 2010
    Level 1 (4 points)
    Jun 18, 2015 4:38 PM in response to AJ 2010

    I hope I'm not the only one who filed a bug report but I did just get this request for more information. Unfortunately it was such a problem that I reverted back and don't think I'll be able to comply. I wanted to put this here just in case someone else runs into it and is able to submit the information below.

     

    We need more information to investigate this issue.

     

    Please provide the following:

     

    Collecting a Packet Trace

    Packet trace is collected by using Terminal.app. To start the trace, you will type a command, followed by the Return key.

    Type: sudo tcpdump -i iptap -s 0 -w ~/Desktop/DumpFile.pcap

    Note: "-s 0" includes a zero, not the letter O.

    When prompted for a password, enter the one for your administrator account.

    Control-C stops collecting the packet trace

     

    Instructions

    Note down the current date and time

    Start a packet capture in Terminal.app

    Mount the server using SMB and repeat the operation when its slow and let about 3 minutes of capture take place

    Stop the packet capture in Terminal.app

    Note down all the names of the files/folders that are involved

    Send us the packet capture, date and time, names of the involved items and /var/log/system.log

     

    Repeat the above, but this time when the server is running fast so we can compare the two data sets against each other.

     

    Please provide your response or results by updating your bug report.

     

    Please compress any bundled files (e.g. nested folders) prior to uploading.

  • by Peter Goldman1,

    Peter Goldman1 Peter Goldman1 Jun 18, 2015 5:18 PM in response to AJ 2010
    Level 1 (24 points)
    Jun 18, 2015 5:18 PM in response to AJ 2010

    Thanks for doing this AJ.

     

    Sadly, I won't be able to comply either.

     

    We've reverted 4 separate networks back to AFP now... same exact server, no other changes... working perfectly.

     

    These are production environments... so unable to recreate this.  Which is a bummer.  We _really_ need this to work....

     

    What else can we do to add fuel to the fire?

  • by AJ 2010,

    AJ 2010 AJ 2010 Jun 18, 2015 5:21 PM in response to Peter Goldman1
    Level 1 (4 points)
    Jun 18, 2015 5:21 PM in response to Peter Goldman1

    Since Logsi had a Wireshark capture I'm really hoping he is still getting notified of replies to this topic so he can create a bug report and submit his capture session.

  • by Logsi,

    Logsi Logsi Jun 19, 2015 1:10 AM in response to AJ 2010
    Level 1 (5 points)
    Jun 19, 2015 1:10 AM in response to AJ 2010

    Since using wireshark to isolate and diagnose which client is causing the problem it's become quite a quick problem to fix. I can just monitor the finders responsiveness or a colleague will say Office Excel is beach balling and i'll log in and disconnect the troublesome client.

     

    I'm not sure if it's a coincidence of either me disabling spotlight indexing the shared drive on client computers or work slowing down here at the moment but the problem doesn't seem to have been as prevalent in the last week or so. Next time it does it I will take a three minute tcp dump and then send a bug report to apple. But if you are having problems give wireshark a go and see if there is anything awry there. If there is no network devices detected use the following command;

     

    sudo chmod 644 /dev/bpf*

     

    Just as an FYI i'm using a Mac Mini server and the shared drive is a LaCie 5big connected via thunderbolt.

     

    I wonder if the change back to mDNSresponder will change the matter come 10.10.4?

  • by Logsi,

    Logsi Logsi Jun 24, 2015 3:29 AM in response to Logsi
    Level 1 (5 points)
    Jun 24, 2015 3:29 AM in response to Logsi

    Have had this problem a lot this morning and when checking the console of the client computers I've noticed a strange log that says this;

     

    Screen-Shot-2015-06-24-at-10.32.47-am.jpg

     

    Quicklook running out of memory!?

  • by AJ 2010,

    AJ 2010 AJ 2010 Jul 1, 2015 2:29 PM in response to Logsi
    Level 1 (4 points)
    Jul 1, 2015 2:29 PM in response to Logsi

    I wonder if 10.10.4 addresses this specific issue. Their release notes are extremely vague. Does it include a fix for extremely slow SMB connections? Who knows. About the OS X Yosemite v10.10.4 Update - Apple Support

     

    This update:

    • Improves networking reliability
  • by LTOguy,

    LTOguy LTOguy Jul 2, 2015 9:45 AM in response to AJ 2010
    Level 1 (0 points)
    Jul 2, 2015 9:45 AM in response to AJ 2010

    Wondering the same thing.  For now I've reverted to AFP and all is well.  Researching using NFS shares via NFSmanager.

     

    Cheers,

    Ernesto

  • by Logsi,

    Logsi Logsi Jul 15, 2015 8:23 AM in response to Logsi
    Level 1 (5 points)
    Jul 15, 2015 8:23 AM in response to Logsi

    Still happening with 10.10.4....

     

    I don't want to have to go back to AFP as that has a different host of problems...

Previous Page 2 of 4 last Next