Page 1 of 3 123 LastLast
Results 1 to 10 of 26
  1. #1
    Untanglit
    Join Date
    Jan 2018
    Posts
    15

    Default Odd dropouts at times

    I had a first gen mobile i5 running fine for years but after upgrading to gigabit found it maxed out at a few hundred megabit, so I upgraded (and reinstalled) with a new Intel server NIC on an HP EliteDesk 800 G1 SFF with an i5-4590, 8GB RAM and a 256GB SSD.

    Internet is AT&T fiber with their router placed in bridge mode. Untangle handles DHCP, routing, etc. WiFi is Eero mesh placed in bridge mode.

    I’m using this at home with around 50 IP devices. Memory is only around 30% used and disk 6% used. CPU load as expected is near zero all the time.

    Tonight I was streaming a movie and then suddenly it froze as did everything else connected to the internet.

    Tried to connect to the Untangle browser from my iPad and it just hung.

    I went down to a wired PC and tried to ping the server and it timed out 4 out of 5 times and returned a ping with 600+ ms the 5th time.

    After 5 minutes it all just came back before I could do anything like examine the terminal itself. This happens once in a while and I’m not sure if it’s an AT&T issue, Untangle (configuration) or hardware. Again, 99.9% of the time it’s just solid. Uptime right now is over 13 days.

    See below from 8:05 to 8:11. Any ideas?

    4D4F984C-CF7E-48A9-9BFA-2A04BA63DDD9.png
    Last edited by Markluke; 06-24-2019 at 05:55 PM.

  2. #2
    Untangle Ninja jcoehoorn's Avatar
    Join Date
    Mar 2010
    Location
    York, NE
    Posts
    1,964

    Default

    I had issues with dropouts several years back, and again a few years before that. In one case, my graph looked much like yours... but I had a corresponding Load graph that would spike during the same interval. I tracked down the source of the load, and the problem went away. In the other case, the graph did not look like yours. Instead of a flat line, there would be nothing at all; just a gap where the traffic would have been. This was a hardware issue. The gaps were where the server was rebooting, or the UVM was restarting.

    I have seen three other situations that have caused graphs like yours: genuine ISP outages, misconfigured network interfaces, and that one time a student bridged his wifi and lan connections on his laptop on the one link where I somehow didn't have STP protection, thus creating a routing loop in my LAN.
    Five time Microsoft ASP.Net MVP managing a Lenovo RD330 / E5-2420 / 16GB with Untangle 16.5.2 to protect a 1Gbps fiber link for ~450 residential college students and associated staff and faculty

  3. #3
    Untangle Ninja sky-knight's Avatar
    Join Date
    Apr 2008
    Location
    Phoenix, AZ
    Posts
    26,542

    Default

    This is a buckle up issue... because tracking it down is HARD!
    Rob Sandling, BS:SWE, MCP
    NexgenAppliances.com
    Phone: 866-794-8879 x201
    Email: support@nexgenappliances.com

  4. #4
    Untanglit
    Join Date
    Jan 2018
    Posts
    15

    Default

    Thanks - I don’t think it’s a NIC misconfiguration as works flawlessly the vast majority of the time. CPU load didn’t spike (would be hard as the i5-4590 is massive overkill for my needs, just use this system as the bus speed is faster for gigabit.)

    Good to know that if rebooting the line would be broken. The system was up, just non-responsive. I would have thought that if there was an ISP issue the system would respond fine, just show no throughput?

  5. #5
    Untangle Ninja sky-knight's Avatar
    Join Date
    Apr 2008
    Location
    Phoenix, AZ
    Posts
    26,542

    Default

    That's just it, was the system unresponsive? If so, which part was unresponsive?

    When this stuff happens, I get on the physical console, because if I can get into the admin UI there, but not over the LAN, then I know I have something up on the LAN side. It's not uncommon for a virus, or a badly configured torrent client to be able to DOS even a gigabit NIC right off the LAN. These issues are easier on larger networks because you get reports of issues isolated to a specific VLAN.

    But that's the sort of thing you're looking at, is it External? Is it Internal? Is Untangle just locking up and restarting? That last one usually is bad RAM by the way...

    Again physical console while its broken, because if you can't even move the mouse, or the num lock on the keyboard won't turn on / off... that's an interrupt control hard lock, that's a dead giveaway for a hardware problem. Bad RAM, or bad disk will do that. The former can even manifest as the latter...
    Rob Sandling, BS:SWE, MCP
    NexgenAppliances.com
    Phone: 866-794-8879 x201
    Email: support@nexgenappliances.com

  6. #6
    Untanglit
    Join Date
    Jan 2018
    Posts
    15

    Default

    Quote Originally Posted by sky-knight View Post
    That's just it, was the system unresponsive? If so, which part was unresponsive?

    When this stuff happens, I get on the physical console, because if I can get into the admin UI there, but not over the LAN, then I know I have something up on the LAN side. It's not uncommon for a virus, or a badly configured torrent client to be able to DOS even a gigabit NIC right off the LAN. These issues are easier on larger networks because you get reports of issues isolated to a specific VLAN.

    But that's the sort of thing you're looking at, is it External? Is it Internal? Is Untangle just locking up and restarting? That last one usually is bad RAM by the way...

    Again physical console while its broken, because if you can't even move the mouse, or the num lock on the keyboard won't turn on / off... that's an interrupt control hard lock, that's a dead giveaway for a hardware problem. Bad RAM, or bad disk will do that. The former can even manifest as the latter...
    Good advice - I had been running my UT server headless on a shelf so by the time I was hooking up a monitor and keyboard it had resolved itself. Will get it ready for the next time.

    The web client was what was not responsive, both wireless from my iPad and wired from a PC.

    Itís possible I have marginal RAM so when the rest of the family is out Iíll probably take it offline and run Memtest.

    I also had IDS and Spam Filter loaded and running, and realized neither is probably doing anything so I turned them off. For DNS am using Googleís .8 and .4.

  7. #7
    Untangle Ninja sky-knight's Avatar
    Join Date
    Apr 2008
    Location
    Phoenix, AZ
    Posts
    26,542

    Default

    Are your wireless and wired networks connected to the same Untangle physical NIC?
    Rob Sandling, BS:SWE, MCP
    NexgenAppliances.com
    Phone: 866-794-8879 x201
    Email: support@nexgenappliances.com

  8. #8
    Untanglit
    Join Date
    Jan 2018
    Posts
    15

    Default

    Quote Originally Posted by sky-knight View Post
    Are your wireless and wired networks connected to the same Untangle physical NIC?
    Yes - wireless connected to a switch which goes to the Untangle LAN NIC port.

  9. #9
    Untangle Ninja sky-knight's Avatar
    Join Date
    Apr 2008
    Location
    Phoenix, AZ
    Posts
    26,542

    Default

    Quote Originally Posted by Markluke View Post
    Yes - wireless connected to a switch which goes to the Untangle LAN NIC port.
    Ahh, sadly that means connecting via wire or wireless tests nothing. For all we know the switch is going bad... something on that switch could be DOS'ing the LAN port off the network... Basically, if the wire doesn't work the wireless won't either.
    Rob Sandling, BS:SWE, MCP
    NexgenAppliances.com
    Phone: 866-794-8879 x201
    Email: support@nexgenappliances.com

  10. #10
    Untanglit
    Join Date
    Jan 2018
    Posts
    15

    Default

    Quote Originally Posted by sky-knight View Post
    Ahh, sadly that means connecting via wire or wireless tests nothing. For all we know the switch is going bad... something on that switch could be DOS'ing the LAN port off the network... Basically, if the wire doesn't work the wireless won't either.
    Yup, all possible. Only way to know more will be for me to access the console if it happens again.

    Is there any benefit to having wireless on a separate NIC? Would still want all to be on the same subnet.

Page 1 of 3 123 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

SEO by vBSEO 3.6.0 PL2