OBPG website going down every day?

[benj]

Hi

The last three days we've noticed that the OBPG website and data servers go down in the afternoon (~1600 UTC) and remains down until around 1000-1100 UTC the next day (give or take a bit, at both ends of that range). The second day of this approximately coincided with your outage for your network maintenance, at least at the start, but the other days obviously not (and I had to wait for the website to come back up before I could post this). It doesn't appear to be just us - traceroute gets as far as NASA's gateway (see below) and checking oceancolor.gsfc.nasa.gov at https://www.uptrends.com/tools/uptime says it's down everywhere. (EDIT: OK, Uptrends says you're down anyway, so maybe it's being fooled by blocking pings, I don't know. The traceroute still says it's not us though!)

Any news on this?

Thanks
Ben

Traceroute:

benj@pmpc1384-bash:~<1001>traceroute oceancolor.gsfc.nasa.gov
traceroute to oceancolor.gsfc.nasa.gov (xx.xxx.xx.xx), 30 hops max, 60 byte packets
1  xx.xxx.xx.xx (xx.xxx.xx.xx)  15.908 ms  15.988 ms  16.061 ms
2  pmfw1.npm.ac.uk (xx.xxx.xx.xx)  0.710 ms  0.821 ms  1.031 ms
3  uop-outbound-rtr.nerc-pml.ac.uk (xx.xxx.xx.xx)  1.567 ms  1.570 ms  1.560 ms <-- Our gateway
4  ge-1-1-2.plymup-rbr1.ja.net (xx.xxx.xx.xx)  1.770 ms  1.765 ms  1.764 ms <-- UK JANET router, so we're getting out
[snipped]
11  et-11-3-0-1275.clpk-core.maxgigapop.net (xx.xxx.xx.xx)  82.853 ms  82.861 ms  82.764 ms
12  xx.xxx.xx.xx (xx.xxx.xx.xx)  84.810 ms  84.858 ms  84.841 ms
13  rtr-s28-hecn-test.sci.gsfc.nasa.gov (xx.xxx.xx.xx)  83.802 ms  83.200 ms  83.395 ms
14  * * *
15  * * *
[snipped, all the same]


[pfsmith]

Ben,
Please send us your IP address as you are seen from the outside. That would allow us to check for network blocks.  The output from https://www.whatismyip.com would be fine. I checked the uptime tool at Uptrends this morning.  Both https://oceancolor.gsfc.nasa.gov and https://oceancolor.sci.gsfc.nasa.gov show as available from all their test points. Not sure why it was appearing down for your check other than possibly using http instead of https. We're strictly https protocol now, so any tests using http will fail.

Also not sure why you are seeing outages three days in a row, other than temporary network blocks.  The downtime for our maintenance day was only about 5 minutes.

Paul

[stgo]

Hi Paul,

Ben asked me to take over on this. The IP is xx.xxx.xx.xx, I'll make sure we test against the https link next time this happens.

Thanks for looking into it.

Stephen

[pfsmith]

Glad to help.  Checking the logs, I found that blocks were triggered on 8/10 and 8/11.  Nothing since. I'll check to see what might have triggered those, but it looks like you are in the clear at the present time.

Paul