question

I am also seeing this issue. I am seeing it consistently with a file a little over 6GB in size. I am on a fiber connection, so the total upload takes about 2.5 minutes for me. I don't think it's a "time to upload" issue since before I had fiber, I had smaller files take MUCH longer and they were successful.

Hi there, We haven't seen any issues related to this from our side. If you could provide any more information on the issue, that would be greatly appreciated. For example, when did any of these errors happen? What was the call doing, and how big was the file? Can we see the request/response, what makes you think it could be related to the load balancer, and is it repeatable, etc etc. Also, is this still an issue after reboot, and/or does it persist with the same file on different accounts or computers? Please provide as much information about the issue as possible, and we will be able to investigate further. Thanks! Jamie

Hi there, We haven't seen any issues related to this from our side. If you could provide any more information on the issue, that would be greatly appreciated. For example, when did any of these errors happen? What was the call doing, and how big was the file? Can we see the request/response, what makes you think it could be related to the load balancer, and is it repeatable, etc etc. Also, is this still an issue after reboot, and/or does it persist with the same file on different accounts or computers? Please provide as much information about the issue as possible, and we will be able to investigate further. Thanks! Jamie

I had this from my user > I have a problem with transmissions from Germany to the USA or the response time. > > If I upload a file from my server in Germany, it is uploaded without any problems. At 100%, but this stops and gives me after 1 minute of this error: "2016/01/31 03:44:11 11.01.2016_12-00/MySQL/Backup.tar.enc: Failed to copy: HTTP code 408: "408 REQUEST_TIMEOUT“. This was doing an upload via the API (a POST request). As for why I think it might be the load balancer, a bit of google searching indicated that "408 REQUEST_TIMEOUT“ was a common error from the amazon load balancers. I'll see if I can get more info for you,

Hi there, Thanks for the extra information. We will take a look, but unfortunately, it's still not really enough for us to properly investigate. We would still need something similar to answers to the questions I asked previously in the thread. If you could try and get this information, we will be happy to investigate further for you. Thanks, Jamie

What you report seems to be a long-standing issue for large files starting at about 10GB (or maybe some gigs less). See https://github.com/yadayada/acd_cli/issues/8. It's either a 408 or 504 status code. The file usually shows up when getting a children listing of the parent folder a few minutes later. People have speculated that the hashing takes too long for large files, so a timeout is triggered. Slightly off topic: I, personally, also have also encountered at least one 408 timeout when uploading a relatively small file. It just vanished. @Jamie C'mon. This has been discussed here before.

I have some logs for you to look at. These are of 504 timeout errors. > We haven't seen any issues related to this from our side. If you could provide any more information on the issue, that would be greatly appreciated. For example, when did any of these errors happen? See the log below. > What was the call doing Uploading a file tocontent-na.drive.amazonaws.com/cdproxy/nodes?suppress=deduplication > and how big was the file? large files >20gb ish. > Can we see the request/response See below. Note that these are headers only, but the log indicates the body for the 504 empty was empty. > what makes you think it could be related to the load balancer A bit of googling suggests this might be the cause, but that is only a guess http://docs.aws.amazon.com/ElasticLoadBalancing/latest/DeveloperGuide/ts-elb-error-message.html#ts-elb-errorcodes-http504 > and is it repeatable Not every time > Also, is this still an issue after reboot Yes and with multiple users / computers / internet providers > and/or does it persist with the same file on different accounts or computers? The files get uploaded eventually. ------------------- Here is an example of request and response 2016/05/04 13:07:45 >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> 2016/05/04 13:07:45 HTTP REQUEST 2016/05/04 13:07:45 POST /cdproxy/nodes?suppress=deduplication HTTP/1.1 Host: content-na.drive.amazonaws.com User-Agent: rclone/v1.29 Content-Length: 0 Content-Type: multipart/form-data; boundary=74fe85afefb8fdaee039fd717f934d9232d5608670c180150ae25605cbeb Accept-Encoding: gzip 2016/05/04 13:07:45 >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> and the response (note empty body and complete lack of headers). 2016/05/04 21:57:02 <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< 2016/05/04 21:57:02 HTTP RESPONSE 2016/05/04 21:57:02 HTTP/1.1 504 GATEWAY_TIMEOUT 2016/05/04 21:57:02 <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< 2016/05/04 21:57:02 pacer: Rate limited, sleeping for 10.660772719s (5 consecutive low level retries) 2016/05/04 21:57:02 pacer: low level retry 1/1 2016/05/04 21:57:02 backupFrom.2016-05-04/bootDiskbakSPLITah.2016-05-04: Received error: HTTP code 504: "504 GATEWAY_TIMEOUT", no response body - low level retry 1/10 Here is a successful response which may be helpful and gives you the time skew between times logged and server time. 2016/05/04 13:07:45 <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< 2016/05/04 13:07:45 HTTP RESPONSE 2016/05/04 13:07:45 HTTP/1.1 201 Created Connection: close Cache-Control: no-store Connection: keep-alive Content-Type: application/json Date: Wed, 04 May 2016 18:07:45 GMT Etag: 6uWh6VZIGJA Location: https://cdws.us-east-1.amazonaws.com/drive/v1//nodes/1jRmFrybTi-akCYRU3IWVA Pragma: no-cache Server: Server Vary: User-Agent X-Amzn-Requestid: 1a0d6755-1223-11e6-9bcf-c12099a86f3c 2016/05/04 13:07:45 <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< Please find more logs of the error here: https://github.com/ncw/rclone/files/250454/log.txt