As many of you have experienced and started reporting at the end of last week, we can confirm on behalf of the Climate Data Store team that there is an ongoing technical issue on the Climate Data Store (CDS) affecting all data retrievals and resulting in significantly increased queueing times.
We are very sorry for the inconvenience this is causing our users.
In the meantime, please be assured that the Climate Data Store team is making every effort to get this issue fixed as soon as possible.
We will post an update as soon as we will have details to share.
We thank you for your patience.
ECMWF Support & CDS team
3 Likes
Hello Anabelle, thanks for letting us know.
Do you have any updates here?
Hi all,
We are pleased to let you know that the Climate Data Store is back!
The Climate Data Store team fixed the issue last night and systems are fully operational again (queues might still be a bit longer than usual while the backlog is being cleared).
Our apologies again for the inconvenience this may have caused.
ECMWF Support & CDS team
2 Likes
Dear all,
Thank you very much for the updates.
For me is still not possible to download ERA5-Land data through the API due to errors during the transfer.
Best,
Diego
Hello,
I still have the same problem: I’ve been in the queue for one day. I am a student, and I need to download the data for my master’s thesis. Is there any solution?
Hello,
I still get a long queue time, don’t know how long will it take to complete the accumulated requests or it is beacuse the issue is not completely fixed yet.
Regards,
The ERA5 queue seems back to normal,
But ERA5 LAND is still pretty much unusable.
Any chance of a fix?
Hello,
I’m also requesting for a fix about ERA5-Land product, the limit of 20 active processes vs more than 1500 queues in simply unrealistic. When does it will be increased?
The Climate Data Store team is investigating an issue with ERA5-Land. We have no further information to share at this time.
Thank you for your patience.
ECMWF Support
Hi! Most of the CAMS Global Composition Forecasts data has been unavailable for the past week. I always get “The tape on which the data reside is unavailable” whenever I try to request a data item. When will this problem be resolved? Thank you!
I still cant download more than 1 year at time even though it was said that june it would be working full capacity. For the past weeks I have been downloading data monthly since 1961 and thats time-consuming. When are we going to be able to download large amount of data at once?
The CDS performance with respect to ERA5-Land data download is linked to users’ behaviour. Some users are making very large requests of ERA5-Land data in NetCDF file format impacting the system memory and therefore degrading the CDS performance for all users.
We urge users to respect the current rules in place for downloading all ERA5 data, including ERA5-Land, i.e. no more than 6000 items at a time, no more than one month of data at a time, preferably in GRIB native format.
Should the issue persists, the Climate Data Store administrator will cancel large data requests and may have to implement further restrictions to allow fair data access to everyone.
Thank you for your cooperation.
ECMWF Support
on behalf of the Climate Data Store team
Hi, when is ERA5 LAND daily precipitation CAN BE PROCURED WITHOUT DISRUPTION. I am not able to download data even for a small area which is badly disrupting my work.
I understand the rules and I am not going to attempt any large data request but I should say that not allowing large data requests impact climatological research by making users have to request over 30 years of data one month at a time, which clearly means wasting days on end to acquire those important data. I just want to know if , by anychance, in the future, large data request will be available again. Took me over 2 weeks of downloading each month at a time and likely to take me weeks to finish my scripts to run those data.Like said by many users, we understand the rules, it is just that many of us have time schedules to follow and not having access to important data means delaying works and TCCs. In regards to rules, Im not fgoing to attempt anyfurther large data requests and I havent even tried large data requests past month, only one month at a time (though I find it to be time consuming and not practical).
In my experience downloading data in GRIB format is faster and more reliable. I assume data is stored in GRIB files originally and the conversion to NetCDF requires significant resources as NetCDF is not designed to “stream” data.
If you only need a couple of locations of data instead of the entire grid, feel free to download them via my open source project Open-Meteo. There is also an open-data AWS distribution with a tutorial to self-host ERA5 as an API using Docker.
This whole situation is quite unsatisfying. We are working in a relatively small area (I think about 300 ERA5 grid cells) and have been struggling to download a historical ERA5 dataset with a limited number of variables for months now. We finally managed to download data for 2000-2020 and are now struggling to get 1990-1999. We also tried OpenMeteo, but the number of API calls for the free tier is very limited, and not all parameters are available or delivered in the same form, so we gave upo on this as well. We already invested much time in just downloading data, our project is delayed and it is hard to explain to the funding agency that we are struggling so much to download a relatievly small dataset. Especially given that the CDS data store is a core Copernicus service. It is also remarkable that this issue is marked as FIXED when it is not. I have been working with CDS data for quite a while and everything was running very smoothly before this upgrade…