Keep getting Connection time out and broken pipe while transfering large folder

  • Ok here is what I having issue with.

    System seems to be fine for the most part, added drives x2-8TB Stripped, FS-BTRFS- SMB share.

    Been testing it out, loaded it up with 3.7TB of data from many different machines, Linux & Windows clients, made a 27GB Test folder.

    Everything seems to be fine other than a few hiccups with too long files names and a couple of time out errors doing large transfers to or from the OMV share.

    Biggest problem now is I get error-"Time out Broken Pipe copying file from OMV.

    Random file if I re-try it goes a few more files and does it again with another file referenced, this is in the processing file window.

    Then it will transfer some and do the same, stall with the Time Out . It will eventually transfer files, if I click the retry or skip, a little clunky but will usually start and finish a transfer, especially if I only use small folder or or 1 large file at a time.

    The same to or from both Win & Linux boxes to ro from TrueNAS NAS.

    Almost seems needs a fdsk, scrub or check disk, like it hangs up doing anything large.

    Machine is the #1 in signature, any other info needed let me know.

    Is there something I can do from CLI to check or log to look at?

    4 Mal editiert, zuletzt von Dudesplace ()

    • Offizieller Beitrag

    Biggest problem now is I get error-"Time out Broken Pipe copying file from OMV.

    This is missing some details. For example, the following suggests that you're pulling "from OMV". What transfer package(s) are you using and what are you pulling to? Client pulls of a large batch of files, using copy and paste, tend to be problematic. If Client pull transfers are needed, you might consider installing an rsync package on your clients.


    added drives x2-8TB Stripped, FS-BTRFS-

    2 each 8TB "stripped" disks? Are you running RAID0?


    ________________________________________________________________________

    What you're describing could be caused by network equipment such as a marginal Ethernet switch or a marginal NIC, as a couple examples. If you're using wireless, interference might be a factor. If you're using a 1GB network, transferring 3.7TB would take a good bit of time which would accentuate an existing issue. Is this problem new or did it always exist?

    Almost seems needs a fdsk, scrub or check disk, like it hangs up doing anything large.

    Maybe there is a problem. Have you checked SMART stat's on the disks?

  • Ok I'll try answer most of these.

    No on RAID0, it's running 2 - 8TB disks striped BTRFS

    SMART shows all good nice low temps as well.

    This problem is just notice, new build so it's first real world stress test.

    Was going to do some hardware changes to other NAS's and this box was donated, looked to make a nice small NAS so I

    polished it up Max RAM and speed, 2-8TB drives and NVME boot.

    Question I have mutable Linux and Win boxes on network with some of the shares mounting.

    Would that effect the speed even idle?

    • Offizieller Beitrag

    No on RAID0, it's running 2 - 8TB disks striped BTRFS

    Two disks would be RAID1. That's not striped.

    Question I have mutable Linux and Win boxes on network with some of the shares mounting.
    Would that effect the speed even idle?

    I don't understand the question. Do you have clients with network shares.?


    You still haven't said if you're using wireless.

    and this box was donated, looked to make a nice small NAS so I

    polished it up Max RAM and speed, 2-8TB drives and NVME boot.

    Perhaps there's a reason why this box was "donated". The hardware may have a problem. Perhaps the hardware should be tested with Stress Linux and memtest86 before building it as a server.

  • bJZujbc


    Network is 2.5GB wired ethernet = Nic-Switch and router between other 2.5GB boxes.

    Will do stress test again but all seemed fine, and it's pretty snappy, just seems somethings amiss transfer wise.

    I did not test mem but I will.

    I'll keep digging to find something.

  • "A person who does precision guesswork based on unreliable information from those who are unsure."


    Like your "Tech support" but I thought you were describing the government till I scrolled up and saw caption.


    Stress testing and memtest comming.

    • Offizieller Beitrag

    Like your "Tech support" but I thought you were describing the government till I scrolled up and saw caption.

    Well,,,, :) , like most around here, I've been surprised many times. There are numerous "X" factors. It's difficult to look into the crystal ball and see the problem.

    Usually, the issues that end up here are either very simple (user doc's would answer many questions) or they're 1 in 1000. The latter can be tricky.
    _____________________________________________________________________________

    BTW: You're running RAID0 and that is striped. That's how you're getting the total storage from two disks, for a total of 14.5TB. If it was a mirror (RAID1) the array would be 7.28TB. That's a risky setup. If anything happens to either disk, all data is lost. I'm guessing you have full backup.

  • Yes fully aware of no redundancy with striped RAID0, it is a test bench, backup to a backup to a backup, but thanks.

    Nothing on it is critical, first using OMV so just feeling it out, I like it, just sorting out quirks.

    BTW all stress and mem tests are good, ran several memory tests. The Dell bios check one and the mem test link provided.

    Somehow during the coarse of that though, I broke my boot, so I had to reinstalled OMV.

    Just an FYI my network is an Orbi RBR850 with 3 satellites, with everything after router is connected 2.5GB network switch and NIC's.

    Except a few 2.5Gb cameras, smarthome hub & devices, phones, iPads, Laptops, TV's etc.

    I will average 30 or so devices online at a time, with never as much as a hiccup on connection or speed between devices.

    Thanks will keep digging and post results.

  • So seems that after a reinstall things are looking much better. However, unfortunately, I did not find anything that stood out that could have been the problem. Just wondering the possibility that maybe the install of the new 2.5GB network card did not install a driver maybe.. not sure.

    But it was doing about the same with the onboard 1GB NIC so don't know. But transfer speeds seem to be pretty consistent average at about

    65 - 80MB/s, dependent on what type a file is transferring, which OS and speed NIC it's going to or from. Many smaller files defiantly slow down.

    But as I think I stated previously, my test folder is a mix of as many types and numbers of files as possible, to make it a realistic test.

    230GB and 17,170 files. But defiantly not hanging and stalling like it was, I can live with that.

    I'l mark the tread as resolved unless I find something else.

  • Dudesplace

    Hat das Label gelöst hinzugefügt.

Jetzt mitmachen!

Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!