Author Topic: Trying to diagnose a laggy server  (Read 1846 times)

sk89q

  • Global Moderator
  • Autococker
  • Posts: 1049
Trying to diagnose a laggy server
« on: March 27, 2009, 12:45:34 AM »
I've had this problem with my server for a while now. It's impossible to run a game server because there is a lot of packet loss. However, that was back then, when we were on the "cheap bandwidth," but now we're redundant via seven ISPs on a 100 mbit pipe, and we're still having the problem. Can anyone care to tell me what I can do to diagnose it? I suspect it's a problem with the data center, and I *really* wouldn't be surprised if it was, but if I want to ask them about it, I would need something to show.

KnacK

  • Global Moderator
  • Autococker
  • Posts: 3039
Re: Trying to diagnose a laggy server
« Reply #1 on: March 27, 2009, 04:20:03 AM »
Describe the server: OS, hardware (mem, cpu, RAID ?)

Have you determined if its network lag or actually caused by the hardware?

sk89q

  • Global Moderator
  • Autococker
  • Posts: 1049
Re: Trying to diagnose a laggy server
« Reply #2 on: March 27, 2009, 09:07:47 AM »
FreeBSD 6.2-RELEASE
CPU: Dual Core AMD Opteron(tm) Processor 275 (2192.82-MHz K8-class CPU)
  Origin = "AuthenticAMD"  Id = 0x20f12  Stepping = 2
  Features=0x178bfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,MMX,FXSR,SSE,SSE2,HTT>
  Features2=0x1<SSE3>
  AMD Features=0xe2500800<SYSCALL,NX,MMX+,FFXSR,LM,3DNow+,3DNow>
  AMD Features2=0x3<LAHF,CMP>
  Cores per package: 2
CPU: Dual Core AMD Opteron(tm) Processor 275 (2192.82-MHz K8-class CPU)
  Origin = "AuthenticAMD"  Id = 0x20f12  Stepping = 2
  Features=0x178bfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,MMX,FXSR,SSE,SSE2,HTT>
  Features2=0x1<SSE3>
  AMD Features=0xe2500800<SYSCALL,NX,MMX+,FFXSR,LM,3DNow+,3DNow>
  AMD Features2=0x3<LAHF,CMP>
  Cores per package: 2
real memory  = 4160684032 (3967 MB)
avail memory = 4016058368 (3830 MB)
em0: <Intel(R) PRO/1000 Network Connection Version - 6.2.9>
em1: <Intel(R) PRO/1000 Network Connection Version - 6.2.9>
ad4: 305245MB <Seagate ST3320620AS 3.AAK> at ata2-master SATA150
da0: <AMCC 9550SX-4LP DISK 3.08> Fixed Direct Access SCSI-5 device

I haven't been able to determine anything.

KnacK

  • Global Moderator
  • Autococker
  • Posts: 3039
Re: Trying to diagnose a laggy server
« Reply #3 on: March 27, 2009, 10:44:12 AM »
you could try running ping from teh command prompt for a boatload of packets and dump that to a text file and see if there is any variance over time.

sk89q

  • Global Moderator
  • Autococker
  • Posts: 1049
Re: Trying to diagnose a laggy server
« Reply #4 on: March 27, 2009, 05:31:25 PM »
I haven't really been able to reproduce actual packet loss by pinging repeatedly. There's flux in the response time, but that could also be due to my connection.

jitspoe

  • Administrator
  • Autococker
  • Posts: 18802
Re: Trying to diagnose a laggy server
« Reply #5 on: March 28, 2009, 10:37:31 PM »
At what rate are you pinging?  Keep in mind the clients typically send 40-80packets/second.

sk89q

  • Global Moderator
  • Autococker
  • Posts: 1049
Re: Trying to diagnose a laggy server
« Reply #6 on: March 31, 2009, 10:28:19 AM »
The packets sent by ping -f are getting dropped by a firewall or something, at least on my end. If I wait a bit, the continuous flood works for a few seconds, but then all the packets start getting dropped.