For the past few hours, like some 2 or 3 weeks ago, I have experienced an awfully low upload speed. A file of 22MB has been uploading for some 25 minutes now, and is still at 70% progress.

I do get this, but it keeps moving without pausing. Occasionally it pauses and times out needing a new retry to the connection. If mine ever pause, it does not start again, but always needs a retry connection.

---

1 Corinthians 9:16 "For though I preach the gospel, I have nothing to glory of: for necessity is laid upon me; yea, woe is unto me, if I preach not the gospel!"
Ephesians 6:18-20, please ;-)
http://tbc-pa.org

Still, the uploads take extremely long.
What's wrong with the server there?

since yesterday, I notice that the upload speed is extremely low and uploads even come to a halt quite often with subsequent retries by the BOINC manager.
Hence, the upload of a finished task can take up to one hour, wheras otherwise it would take only several minutes.

Testing my connection with "Speedtest.net" shows the usual perfect results.
So the problem must be with the GPUGRID server.

the upload problems mentioned above last week still exist.

Could anyone from GPUGRID please look into this?

I get non-existent uploading but then I stop the Surfeasy VPN and hit retry and voom, job done. what is the issue with uploads and VPNs? Do you guys with this issue also run vpns?

Since this morning, I have again been experiencing extremely slow speed to and from the GPUGRID server (which means: uploads + downloads)

Anyone else having the same problem?

I just uploaded one now at 250 kbps. That is the speed limit that I have set in BOINC. Also, it started the download of the next one at the same time, since I have it set for 0% resource share, and it completed the download OK, so it appears that everything is OK here. (That does not answer your question, but it is not the same universal problem as the last time.)

The website is a little slow however; I have some delay in accessing the forum, and previewing and editing my response, for example.

The website is a little slow however; I have some delay in accessing the forum, and previewing and editing my response, for example.

Me, too

The website is a little slow however; I have some delay in accessing the forum, and previewing and editing my response, for example.

today the website sometimes is dramatically slow. It can take up to several minutes :-(
What's wrong?

It comes and goes, and is not so bad at the moment. Maybe a routing problem somewhere? But finding it is another matter.

Is this part of the current problem:

17-Jun-2017 11:50:44 [GPUGRID] Temporarily failed upload of e12s17_e3s21p0f386-PABLO_P01106_2_IDP-0-1-RND7610_0_9: transient HTTP error
17-Jun-2017 11:50:44 [GPUGRID] Backing off 00:00:32 on upload of e12s17_e3s21p0f386-PABLO_P01106_2_IDP-0-1-RND7610_0_9
17-Jun-2017 11:50:45 [---] Project communication failed: attempting access to reference site
17-Jun-2017 11:50:48 [---] Internet access OK - project servers may be temporarily down.
17-Jun-2017 11:51:17 [GPUGRID] Started upload of e12s17_e3s21p0f386-PABLO_P01106_2_IDP-0-1-RND7610_0_9
17-Jun-2017 11:51:23 [GPUGRID] Error reported by file upload server: [e12s17_e3s21p0f386-PABLO_P01106_2_IDP-0-1-RND7610_0_9] locked by file_upload_handler PID=14513
17-Jun-2017 11:51:23 [GPUGRID] Temporarily failed upload of e12s17_e3s21p0f386-PABLO_P01106_2_IDP-0-1-RND7610_0_9: transient upload error

rinse and repeat.

Thanks,

Win

right after my post above I checked again and the upload is in progress 124KBPs.

W

Is this part of the current problem:

locked by file_upload_handler PID=14513
[GPUGRID] Temporarily failed upload of e12s17_e3s21p0f386-PABLO_P01106_2_IDP-0-1-RND7610_0_9: transient upload error

This specific problem was also discussed in this thread:
http://gpugrid.net/forum_thread.php?id=4477
However, there were never any comments from GPUGRID-side :-(

locked by file_upload_handler PID=14513

is one of those things which happens from time to time, and is designed to be handled. Process ID 14513 will run its allotted time span and be terminated: the lock will be released: and another file upload handler will be started to take its place. That copes with the occasional handshake timing errors which interrupt network transmissions.

In opening this forum input window, I got a website timeout error and had to wait a few seconds. Maybe the time is approaching for the server to be given a monthly (or three-monthly, or annual) reboot to clear the stale memory images of things that have gone wrong in the past?

Maybe the time is approaching for the server to be given a monthly (or three-monthly, or annual) reboot to clear the stale memory images of things that have gone wrong in the past?

Let's keep our fingers crossed that someone from GPUGRID will read this within the next weeks or months.

In any case, the problems with the download server still persist - downloading a new task now can take up to 20 minutes (before, with my internet bandwidth, it was a matter of about 15 seconds)

So far, the internet connection to GPUGRID has further deteriorated.

Not only that downloads have been extremely slow within the past weeks, now also uploads take forever.

But, unfortunately enough, I bet that the GPUGRID people haven't even noticed this.
Plus, as we know, they re not reading in the forum here either :-(

So far, the internet connection to GPUGRID has further deteriorated.

Not only that downloads have been extremely slow within the past weeks, now also uploads take forever.

I have also noticed the uploads have gone down in speed, I am getting 1mbps vs the usual 4mbps so the WU takes 4 times longer to upload meaning the next person can't start computing without the previous WU in the chain being uploaded.

... meaning the next person can't start computing without the previous WU in the chain being uploaded.

Don't worry about that. While there are unsent workunits in the queue, everybody can have their next piece of work.