Packet Loss to Crucial.com.au and my sites

[cre8ive]

Hi All,

For most of today my site has been experiencing very poor performance and HTTPS pages have not loaded at all.

If I ping crucial.com.au or my site cuffed.com.au I receive packet loss - over 20%.

I also cannot access any HTTPS pages on my site eg https://www.cuffed.com.au .

Have a look at
http://just-ping.com/index.php?vh=cr....au&c=&s=ping!
and http://just-ping.com/index.php?vh=cr....au&c=&s=ping! to see the ping results from around the world.

I have been in contact with Technical support however they are stating they are unable to replciate the issue and everything is fine. I find this hard to believe as crucial.com.au itself is experiencing issues.

Thanks,

Ben

[ross]

Hi cre8ive,

If you feel Tech Support have not provided you with a valid response, by all means please shoot through your concerns with any relevant Ticket ID's to feedback at crucial.com.au.

[splintex]

Massive loss, ssh intermittent access, server console times out, has been stable ever since I joined, and just this afternoon evening, it is all going pear shaped. Also issues as above with connecting to crucial website.

[splintex]

I do realise this is an old thread, but my issues were identical. Massive packet loss to my server, massive packet loss to crucial.com.au. I could not access the server console through the dashboard (when I could connect) and services running on my box were adversly effected (packetloss sucks eh...).

It did rectify itself after about 45mins, and I could then operate with no issues, but during that time in the breif instants I could access my box, vnstat didn't show any abnormal bandwidth usage, no runaway tasks, normal memory usage, nothing abnormal seemed to be happening on the server, so I can only thing there must have been some issue with the network infrastructure.

Anyone?

Also, when I tunnel firefox through ssh I still get redirected to google.com.kh and geoloc sites report that my IP is in bangladesh... which has never happened until earlier this week (see other post)

-Splint

[Aaron]

Hi Splint,

There were no reported issues over the last few days regarding the issues you mentioned. We have monitoring from a number of monitoring stations globally to monitor packetloss, and nothing was reported over the past few days.

This issue may have been isolated to your service or an issue with your ISP. When the issue comes up again, please provide a ping, traceroute, and WinMTR (100 responses) and we can have a look into why this may be happening.

[George]

Originally Posted by splintex

I do realise this is an old thread, but my issues were identical. Massive packet loss to my server, massive packet loss to crucial.com.au.

-Splint

This is exactly what happened to me last night.

Crucial support could find absolutely no issues with the servers or network, and no packet loss from anywhere.

Also my own personal internet connection was working perfectly with any website OTHER than crucial or my site hosted by crucial.

I've already taken it up with my ISP (Optus) but they're not finding any issues.

Anyone have any ideas at all?

[ross]

Hi Geroge,

As mentioned via Whirlpool, the root cause can be a number of things, such issues are often not easily traced down.

It can be something to do with our network provider (vocus / and others) or somewhere inbetween us and your ISP, or something with your ISP.

There is a number of causes, ranging from a bad route, a faulty piece of hardware, DDOS/DOS, server problems, network configuration fault, the list goes on.

The best way to try and track down intermittent issues such as this is to gather as much information as possible.

Simply providing a traceroute doesn't help much. If you do have the issue again, please provide Tech Support pings, traceroutes, winmtr results. It can help us a great deal when as much information is gathered during the affected period.

[George]

Hi Ross,

As it happens, when the issue arose I logged a support ticket. I did provide details pings and other information as requested to tech support. They agreed there was packet loss but said it was due to "Node 5" (of my supplied traceroute). However this turned out to be false, as we have now discovered that node is a router that is simply non-responsive to ICMP packets.

Back to square one.

Just out of interest, here is some WinMTR data from the past 10 minutes or so. You'll notice I've got some minor packet loss to crucial sites but no others (except the aforementioned unresponsive Node 5):

|------------------------------------------------------------------------------------------|
| WinMTR statistics |
| Host - % | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
| myrouter.home - 0 | 1409 | 1409 | 0 | 0 | 5 | 0 |
| per3.ba.optusnet.com.au - 0 | 1409 | 1409 | 26 | 30 | 326 | 27 |
| per1-ge0-0-447.gw.optusnet.com.au - 0 | 1409 | 1409 | 26 | 27 | 75 | 27 |
| xe-3-1-0.pd11.optus.net.au - 0 | 1409 | 1409 | 26 | 28 | 75 | 27 |
| No response from host - 100 | 281 | 0 | 0 | 0 | 0 | 0 |
| 119.225.39.90 - 0 | 1409 | 1409 | 71 | 73 | 150 | 73 |
| ge-0-0-2.bdr02.mel04.vic.VOCUS.net.au - 0 | 1409 | 1409 | 86 | 89 | 153 | 87 |
| ge-0-0-3.bdr01.mel04.vic.VOCUS.net.au - 0 | 1409 | 1409 | 86 | 89 | 157 | 88 |
| ge-0-1-0.cor01.syd03.nsw.VOCUS.net.au - 0 | 1409 | 1409 | 86 | 88 | 184 | 87 |
| ten-0-1-0.bdr01.syd03.nsw.VOCUS.net.au - 0 | 1409 | 1409 | 82 | 84 | 153 | 83 |
|as38544-2.cust.bdr02.syd03.nsw.VOCUS.net.au - 0 | 1409 | 1409 | 82 | 88 | 458 | 83 |
| 111.118.161.101.reverse.crucialx.net - 1 | 1357 | 1344 | 86 | 87 | 159 | 87 |
| sh2-eqx3-syd.ha-node.net - 2 | 1329 | 1309 | 86 | 87 | 137 | 88 |
|________________________________________________| ______|______|______|______|______|______|
WinMTR v0.92 GPL V2 by Appnor MSP - Fully Managed Hosting & Cloud Provider

[ross]

Hi George,

That winmtr results looks pretty standard.

Ping's to our shared hosting servers will get rate limited this is a feature of the firewall, so some ping packets may get dropped along the way. This is too be expected.

The % column is the key indicator, if that was like 5, 10 or even higher, we would know there was an issue with one of the hops being slow or having issues.