NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
Hi Sean,How much warning (if any) do you plan on giving before switching back to 4.5 on the production server? And will there be a chance to access the test server before the switch since the new server will (presumably) have changed URLs and variable names and Unidata doesn't have any documentation on what those changes are?
BTW, checking the http response status code doesn't always work. For example, if you start downloading a large file before the server is shut down, and the server goes down during the download then you end up with a nice corrupted file.
On 4/2/14 11:00 AM, thredds-request@xxxxxxxxxxxxxxxx wrote:
Hi Don, all, I agree that would be useful. At the moment, these restarts are done manually as needed, and the process is not run on a cron. We hope to get 4.5 on our test server very soon (today), and if the test phase goes well, then we will not need to restart on a regular basis. In the meantime, I would strongly suggest that users check the http response status codes and handle those appropriately. For example, if the response code indicates that the server is temporarily unavailable, have the script sleep for a minute or two and try again (these restart periods should be negligible in duration, so the sleep probably does not need to be more than a minute). Not only will this immediately make scripts more reliable, it's the way any web enabled client or script should approach http communications (best practices kind of thing). For users who want to know about the status codes, please see: http://www.w3.org/Protocols/rfc2616/rfc2616-sec10.html If users need help, we'd be happy to advise them where we can. Remember, we are not operational (the UCAR network is not even considered operational), and users will encounter more than a status code of 200. Cheers, Sean On 4/2/14, 8:34 AM, Don Murray (NOAA Affiliate) wrote:Hi Sean- On 4/1/14 9:26 AM, Sean Arms wrote:2) IDV Users, part 1: Everything should be back to how it was prior to last Friday. However, we will need to restart the TDS a few times each day, and this will result in thredds.ucar.edu being unavailable for short windows of time (things usually bounce within 2-3 minutes). If you notice that a collection has stopped updating, please let us know via email to support-thredds@xxxxxxxxxxxxxxxx.Could you please post the times that you will be restarting the TDS so that we can adjust our automatically updating bundles and cron scripts to not hit the server at those times. Thanks. Don
thredds archives: