Degree Days.net Status Updates

UPDATE: the central service is running again after a major outage in one of their data centres, and, from about 2024-06-25 16:00 GMT onwards, we have again been able to provide the latest data through the Degree Days.net website and API. Apologies again for any inconvenience caused!

ORIGINAL: There’s a major problem with the central service that aggregates weather reports from ICAO and WMO weather stations worldwide. Fresh reports have not been coming through since late on the 24th (GMT time). We are unable to calculate degree days for the 24th until we have the necessary weather reports for that day and at least one report for the 25th. For many stations this means we are still unable to calculate degree days for the 24th. If you are trying to fetch data for a longer period that includes the 24th, you may find that that figures for the 24th are missing. Or, if you are trying to fetch data for the 24th only, you may get a SourceDataCoverage failure (or exception/error depending on the programming language you are using).

Unfortunately there is nothing we can do but wait for the central service to fix this problem. When it does, Degree Days.net will automatically make the latest data available, as it usually does.

The underlying problem is affecting virtually all weather-data providers, as virtually all weather-data providers depend on the weather reports from ICAO and WMO weather stations, and ultimately depend on the central service that aggregates these reports and makes them available. We apologize for any inconvenience caused! Hopefully this issue will be resolved soon (on the 25th).

UPDATE: AWS appear to have fixed many of their issues, and our system appears to have been functioning normally from about 2021-12-07T22:10Z onwards. Apologies again for these problems!

ORIGINAL: Our API runs on Amazon Web Services, which, unfortunately, is having a major outage today. It’s not a complete blackout, but it is causing some calls to our API to fail, and this is having knock-on effects on our website too.

We are basically stuck waiting for Amazon to fix the issues with their services. Many other websites and web services are affected as well, as so many use AWS for hosting. When the AWS status page shows all green ticks, the intermittent errors in our service should stop. Hopefully it won’t be too long!

We apologize sincerely for any inconvenience!

UPDATE: the problem seems to have been resolved from about 2017-01-27 17:00 GMT.  Sorry again for this problem!

ORIGINAL: Unfortunately for many of the weather stations in our system the last day or two of data is not currently available. There was a recent change made to Weather Underground’s system that has meant we are currently unable to access fresh temperature data until we can resolve the problem with them. This is affecting both the Degree Days.net website and API.

We are hoping to have this problem resolved at some point today (January 27th, 2017).

Our sincere apologies for this inconvenience!

12:30 GMT: Looks like it’s working again, so Degree Days.net should be back to normal now.

12:07 GMT: Weather Underground’s system has just become inaccessible. It looks like their whole system is down – the website is saying it’s overloaded. This is preventing our system from accessing the most recent temperature data, affecting both the API and website. Our system will start working again once their system is working fully again.

UPDATE 19:50 GMT: it all seems to be accessible again. We’ll be keeping an eye on it, but for the moment it certainly looks like the problem is over and everything should be back to normal again.

UPDATE 19:48 GMT: it looks like the system is partially up and running again. We’re getting some of the temperature data OK now, and hopefully the rest will be accessible soon too.

19:30ish GMT Something has just gone wrong with Weather Underground’s database of temperature data. We can’t access fresh temperature readings out of their systems; only the historical data that is already stored in our system.

What this means is that, until this problem is fixed (hopefully very soon), for many stations you’ll find the latest degree days (last day or two) are inaccessible. This is affecting the API and website.

UPDATE – 2013-04-08 16:32 GMT – it’s been working fine again for the last 10 minutes, and will hopefully continue to do so. Sorry again for the problem.

The problem described in this post has arisen again, affecting our API and website. See the linked post for the details. Sincere apologies!

UPDATE – 2012-01-06 22:39 GMT – Problem is solved, and everything should now be back to normal. Sorry for any inconvenience!

ORIGINAL NOTE: Our system is currently struggling to fetch temperature data from Weather Underground’s system. This is causing our API to respond slowly, requests for some lower-quality weather stations to fail (the PWS for which we are using average/maximum/minimum data straight out of WU’s system to calculate degree days), and the last day or two of data to be missing for the higher-quality stations (airports and the best PWS for which we are using our own optimized cache of temperature data but for which we still rely on WU’s system to fetch fresh data as it comes in).

All should recover once Weather Underground’s system is functioning fully again.

Update: all has been working again since about 09:45 GMT. The AWS status page shows all the issues as resolved, and our system isn’t experiencing any further problems either.

Original problem report below:

For the last few hours, several of the Amazon Web Services that our system depends on have been playing up. As far as we can tell the faults are intermittent, and are causing slowness and parts of our system to fail some of the time.

If you’re using the API, you may be noticing that the API is responding slower than usual, and you might also be seeing an unusual number of ServiceTemporarilyDown failures. (These are to be expected from time to time, so it is good for your code to automatically retry if it gets them, but we know that many of our customers have not yet built retries into their data fetching, so unfortunately this could cause bigger problems for those customers if e.g. scheduled tasks exit on the first ServiceTemporarilyDown failure.)

The website will be failing intermittently too.

We’re hosted across multiple availability zones which appears to be helping to reduce the impact of these problems. Our system also appears to be coping reasonably well without all the services it uses when fully operational. But ultimately we need AWS to get their system fully functional again before everything will start working again completely as it’s supposed to. The AWS status page has up-to-date information on these issues.

We’re very sorry if this is causing you problems. If you are getting errors, the best we can suggest is that you keep trying as the faults should only be affecting some requests some of the time.