Re: [tlug] Packet loss: How much is normal, and who should I talk to

[Darren Cook <darren@example.com>]

> Here's what mtr is reporting. Note the packet loss numbers in the first
> column:

I just tried (but with report cycles set to 10), and got 0%, up until
10% once inside Rackspace, i.e. the same point it jumps from 17% to 87%
for you:

 17. xe-0-2-0.mpr1.ord5.us.above.  0.0%    10  146.0 152.1 146.0 183.8  13.3
 18. 208.185.125.6.IPYX-076520-ZY  0.0%    10  148.4 148.8 148.4 149.4   0.3
 19. coreb.ord1.rackspace.net      0.0%    10  152.9 153.8 152.8 154.9   0.9
 20. Core3-CoreA.ord1.rackspace.n 10.0%    10  153.0 153.4 153.0 153.7   0.2
 21. aggr601a-3-core3.ord1.racksp 10.0%    10  158.1 158.3 157.8 160.5   0.8
 22. secure.emailsrvr.com         10.0%    10  156.9 154.5 146.4 166.2   5.9


I.e. the first people you need to talk to are Rackspace :-)

I think you should expect 0% packet loss on wired connections; for
wireless it will depend on lots of factors as is unlikely to be 0%. But
wireless would just be the first hop or two.

So I would also be concerned about the usen to ipix 17% loss.

Darren


> 
> /usr/sbin/mtr --report --report-cycles 1000 secure.emailsrvr.com
> 
> HOST: cocoa Loss% Snt Last Avg Best Wrst StDev
> 1. gateway.office.agos.co.jp 0.0% 1000 0.2 0.2 0.1 0.3 0.0
> 2. 124x39x186x17.ap124.ftth.uco 0.0% 1000 0.4 0.4 0.4 9.9 0.8
> 3. 221x112x12x97.ap221.ftth.uco 1.5% 1000 0.7 0.9 0.4 51.9 3.5
> 4. 221x112x21x77.ap221.ftth.uco 0.2% 1000 0.8 0.7 0.6 44.5 2.1
> 5. 221x112x16x53.ap221.ftth.uco 0.0% 1000 0.7 1.5 0.4 27.7 3.2
> 6. 61.122.113.121 1.2% 1000 16.5 23.5 0.7 53.0 9.7
> 7. usen-61x122x114x82.gate01.co 0.1% 1000 1.1 1.2 0.8 22.6 1.7
> 8. as6461.ix.jpix.ad.jp 17.6% 1000 15.7 22.6 1.7 60.9 8.8
> 9. so-7-0-0.mpr1.sjc2.us.above. 17.1% 1000 127.6 141.3 121.6 191.5 7.9
> 10. ge-9-2-9.er2.sjc2.us.above.n 20.6% 1000 124.8 134.3 114.7 231.2 9.7
> 11. xe-1-3-0.cr2.sjc2.us.above.n 17.5% 1000 128.9 143.5 121.2 214.0 11.7
> 12. ae3.cr2.ord2.us.above.net 18.5% 1000 185.7 185.7 160.7 226.4 9.5
> 13. xe-2-1-0.mpr2.ord6.us.above. 19.7% 1000 189.9 188.9 165.4 278.1 11.1
> 14. xe-0-2-0.mpr1.ord5.us.above. 21.8% 1000 184.3 188.0 163.5 268.3 12.6
> 15. 208.185.125.6.IPYX-076520-ZY 19.2% 1000 189.1 195.5 174.7 304.5 8.0
> 16. corea.ord1.rackspace.net 16.9% 1000 180.0 186.3 165.8 214.1 6.8
> 17. Core3-CoreB.ord1.rackspace.n 87.1% 1000 181.9 194.1 177.8 205.5 7.2
> 18. aggr601a-3-core3.ord1.racksp 99.7% 1000 211.2 203.9 189.3 211.3 12.7
> 19. secure.emailsrvr.com 15.5% 1000 177.0 187.0 156.8 214.1 9.1
> 
> By contrast, from home (which is on a different ISP to the one shared by
> the main office and a satellite office that have both reported problems),
> the same thing gives me hardly any packet loss (I think the ??? one is my
> cable router that doesn't accept pings):
> 
> HOST: beech Loss% Snt Last Avg Best Wrst StDev
> 1.|-- 192.168.1.1 0.0% 1000 0.3 0.3 0.3 0.6 0.0
> 2.|-- ??? 100.0 1000 0.0 0.0 0.0 0.0 0.0
> 3.|-- 10.202.113.3 0.0% 1000 9.9 11.6 8.2 42.9 3.1
> 4.|-- 172.25.106.229 0.0% 1000 10.9 12.7 7.8 84.3 5.5
> 5.|-- 10.1.1.249 0.0% 1000 11.0 13.5 9.0 117.6 5.9
> 6.|-- 203-165-18-217.rev.home.n 0.0% 1000 11.3 16.2 9.4 89.3 11.9
> 7.|-- 203-165-18-218.rev.home.n 0.0% 1000 13.8 16.1 9.2 114.8 6.8
> 8.|-- c1-t0-2-0-6.kdd1.nap.home 0.0% 1000 17.4 15.2 10.0 75.2 4.9
> 9.|-- 124.211.14.13 0.0% 1000 16.4 14.5 9.7 77.9 6.5
> 10.|-- otejbb206.int-gw.kddi.ne. 0.0% 1000 13.8 21.4 9.1 118.2 24.2
> 11.|-- pajbb002.int-gw.kddi.ne.j 0.0% 1000 131.8 134.4 126.8 239.9 12.9
> 12.|-- ix-pa9.int-gw.kddi.ne.jp 0.0% 1000 123.5 128.9 119.8 307.0 23.3
> 13.|-- 64.125.12.61 0.0% 1000 123.2 127.5 118.4 220.0 13.8
> 14.|-- xe-3-1-0.cr2.sjc2.us.abov 0.0% 1000 134.6 143.5 131.5 206.3 9.1
> | `|-- 64.125.31.70
> 15.|-- ae1.cr2.lax112.us.above.n 0.0% 1000 140.4 141.7 134.4 224.6 10.4
> 16.|-- ae4.cr2.iah1.us.above.net 0.0% 1000 135.7 172.2 133.3 258.8 18.7
> | `|-- 64.125.30.233
> 17.|-- ae1.cr1.iah1.us.above.net 0.0% 1000 178.0 183.4 175.5 264.3 10.3
> | `|-- 64.125.25.46
> 18.|-- ae3.cr1.dfw2.us.above.net 0.9% 1000 175.1 176.1 167.3 223.8 8.6
> 19.|-- ae4.er1.dfw2.us.above.net 2.4% 1000 172.2 177.0 170.9 227.2 6.4
> 20.|-- 128.177.70.86.IPYX-076520 0.0% 1000 179.0 185.2 178.8 278.3 10.5
> 21.|-- corea.dfw1.rackspace.net 0.1% 1000 186.3 190.9 183.0 262.1 8.9
> 22.|-- core5.dfw1.rackspace.net 0.1% 1000 172.1 175.9 169.4 230.8 4.0
> 23.|-- aggr921a-10-core5.dfw1.ra 0.9% 1000 213.9 213.5 194.8 291.7 7.8
> 24.|-- secure.emailsrvr.com 0.1% 1000 172.3 180.0 169.2 254.6 6.8
> 
> Rackspace then apparently chose that precise moment to outsource their
> support operations to a brick wall, so I'm trying to figure out whether I
> should be arguing with them, arguing with my ISP USEN, fixing something
> myself, or forgetting about network issues and tinkering with anti-virus
> software and mail clients and things instead.
> 
> So I'm wondering:
> - Where I see a fair bit of packet loss before I get to the US, is that
> something exceptionally bad that shows there's already a problem at my end
> / with USEN, or is it par for the course?
> - When it gets to the Rackspace systems and we start getting 99.4% packet
> loss listed, does that pretty much conclusively show there's a problem at
> their end, or could that show up as the result of failures closer to me?
> Alternatively, is it something that could happen with a
> normally-functioning system for whatever reason?
> 
> Any thoughts?
> 
> 
> 


-- 
Darren Cook, Software Researcher/Developer

http://dcook.org/work/ (About me and my work)
http://dcook.org/blogs.html (My blogs and articles)