Failed to connect to 171.64.65.104:80

[Simplex0]

Thank you Joe_H!

I have read now that when you in the servers 'Status' column read the word 'full' it stands for 'full operation' and that everything is ok, maybe 'ok' would have been a better word to use in this case.
'Hi there! What's your status?' "Hi! Currently I'm full"

[bruce]

OK would not be a better word. "Full" indicates that it's configured for either sending or receiving WUs, though the server may not be okay.

For example, see 171.67.108.149
... GPU full DOWN ...

[Simplex0]

According to the information given on the server page the word 'full' in the 'Status' column means "Server is in full operation, everyting is ok" so "ok" would have worked just fine

[kwerboom]

Greetings. I think that server 171.64.65.104:80 is once again sick (or full or rejecting or not accepting). Could you please fix it soon. Thanks.

Code: Select all

19:54:45:WU01:FS01:Upload complete
19:54:45:WU01:FS01:Server responded PLEASE_WAIT (464)
19:54:45:WARNING:WU01:FS01:Failed to send results, will try again later
20:03:48:WU01:FS01:Sending unit results: id:01 state:SEND error:NO_ERROR project:9205 run:49 clone:11 gen:8 core:0x21 unit:0x0000001b664f2dd056fb26b7beb92751
20:03:48:WU01:FS01:Uploading 37.97MiB to 171.64.65.104
20:03:48:WU01:FS01:Connecting to 171.64.65.104:8080
20:03:49:WARNING:WU01:FS01:WorkServer connection failed on port 8080 trying 80
20:03:49:WU01:FS01:Connecting to 171.64.65.104:80
20:03:51:WARNING:WU01:FS01:Exception: Failed to send results to work server: Failed to connect to 171.64.65.104:80: No connection could be made because the target machine actively refused it.
20:03:51:WU01:FS01:Trying to send results to collection server
20:03:51:WU01:FS01:Uploading 37.97MiB to 171.65.103.160
20:03:51:WU01:FS01:Connecting to 171.65.103.160:8080

[DutchForce]

Update from the project owner, Jadeshi... p92xx temporarily offline

by jadeshi » Mon Jul 18, 2016

Hey all,
Projects 9205-9213 will temporarily not be accepting any work requests due to the WS running out of space. I am currently
trying to resolve this and and will update you as soon as this is fixed. Sorry for the inconvenience.

[bruce]

According to the information given on the server page the word 'full' in the 'Status' column means "Server is in full operation, everyting is ok" so "ok" would have worked just fine

No, you missed my point.

You can only understand the Status and Connect columns in conjunction. A server can be configured for Full operation or Accept (half operation) or Standby (intentionally off-line) but even if it's configured for FULL or Accept, it can also completely DOWN, in which case it would not be "OK."

Apparently you didn't look at the example I mentioned.

[_r2w_ben]

Does this means that the servers hard drives is full?
"
Mon Jul 18 09:00:25 PDT 2016 171.64.65.104 vspg14b jadeshi GPU full Reject 0.00 0 0 50541 9913 -1563 0 0 0 - - - - - 0 0 - - 1 - 0 0 WL; WL; 10000, 10000 7.0, 7.0 - 49, 49 64, 64 - - 2, 1 B, B 8080G, 8080G
"

It's hard to tell from the log that the disk is full. Hovering over the i icon at the end of the row for 171.64.65.104 on the server status indicates:

Code: Select all

GB Tot[al]: 50541
GB Av[ailable]: 9913

These numbers appear in the middle of your log snippet. 50541 stays constant as you scroll down the log while the next number drops regularly. Without knowing how the disk space is allocated, 9TB would appear to be sufficient space. If this number had dropped to 0 then it would be more obvious.

[Joe_H]

Those numbers appear to be space totals for the entire vspg14 server, there are at least 5 different VM's hosting projects on that server. If you hover over the "i" icon for the other WS entries for vspg14a-e, the same numbers are shown. Three other WS VM's hosted there are accepting WU's back. My assumption, based on the project managers post that the 171.64.65.104 WS is out of space, is that the storage allocated to that VM was filled.

[Simplex0]

This seams to take some time to solve. running BOINC in the mean time.

[jdmurray]

And same here too. I have a large WU (34820) ready to send and the F@H client cannot connect to 171.64.65.104:80 work server. As an alternate, the collection server 171.65.103.160:8080 receives the upload, but apparently does not give the proper response (PLEASE_WAIT (464)).

I see on the server stats page that 171.64.65.104 is full and rejecting connections, but 171.65.103.160 appears nominal.

Code: Select all

14:30:47:WU00:FS01:Sending unit results: id:00 state:SEND error:NO_ERROR project:9206 run:0 clone:192 gen:41 core:0x21 unit:0x00000069664f2dd056fb26bc320bcd90
14:30:47:WU00:FS01:Uploading 37.24MiB to 171.64.65.104
14:30:47:WU00:FS01:Connecting to 171.64.65.104:8080
14:30:48:WARNING:WU00:FS01:WorkServer connection failed on port 8080 trying 80
14:30:48:WU00:FS01:Connecting to 171.64.65.104:80
14:30:49:WARNING:WU00:FS01:Exception: Failed to send results to work server: Failed to connect to 171.64.65.104:80: No connection could be made because the target machine actively refused it.
14:30:49:WU00:FS01:Trying to send results to collection server
14:30:49:WU00:FS01:Uploading 37.24MiB to 171.65.103.160
14:30:49:WU00:FS01:Connecting to 171.65.103.160:8080
14:30:55:WU00:FS01:Upload 43.47%
14:31:01:WU00:FS01:Upload 97.01%
14:31:01:WU00:FS01:Upload complete
14:31:01:WU00:FS01:Server responded PLEASE_WAIT (464)
14:31:01:WARNING:WU00:FS01:Failed to send results, will try again later

[jdmurray]

My big WU just disappeared from the work queue and this was in the log:

Code: Select all

13:38:00:WARNING:WU00:FS01:Past final deadline 2016-07-19T06:08:38Z, dumping
13:38:07:WU00:FS01:Cleaning up

My machine spent a week chewing on that WU only to have all that work thrown away. Is there a way to specify only smaller WUs are to be downloaded?

[Simplex0]

My big WU just disappeared from the work queue and this was in the log:

Code: Select all

13:38:00:WARNING:WU00:FS01:Past final deadline 2016-07-19T06:08:38Z, dumping
13:38:07:WU00:FS01:Cleaning up

My machine spent a week chewing on that WU only to have all that work thrown away. Is there a way to specify only smaller WUs are to be downloaded?

That's just sucks! I have 2 finished wu that have been waiting to be uploaded for some time now. I have stopped runing folding and just waiting for this 2 to be uploaded before I shut it down, hopefully they will.

[bruce]

I see on the server stats page that 171.64.65.104 is full and rejecting connections

Where do you see that? I see it is in standby (and reject) which means it's intentionally off-line. That agrees with p92xx temporarily offline, which states that the server is being worked on.

Yes, it sucks when a WU expires and is dumped.

The final deadline for those projects is 10 days after downloading. What GPU are you running? How many hours per week do you fold?

[petem]

Won't pile on, but FYI I have 2 WUs that are sitting in limbo as well - same issues - can't connect to 171.64.65.104, failed upload to 171.65.103.160.
Your attention to this issue is greatly appreciated ; )

You probably don't need it, but just in case...

Code: Select all

...
20:40:18:WU01:FS01:Sending unit results: id:01 state:SEND error:NO_ERROR project:9205 run:41 clone:1 gen:61 core:0x21 unit:0x00000083664f2dd056fb26b3f7219a9b
20:40:18:WU01:FS01:Uploading 37.73MiB to 171.64.65.104
20:40:18:WU01:FS01:Connecting to 171.64.65.104:8080
20:40:19:WARNING:WU01:FS01:WorkServer connection failed on port 8080 trying 80
20:40:19:WU01:FS01:Connecting to 171.64.65.104:80
20:40:19:WARNING:WU01:FS01:Exception: Failed to send results to work server: Failed to connect to 171.64.65.104:80: Connection refused
20:40:19:WU01:FS01:Trying to send results to collection server
20:40:19:WU01:FS01:Uploading 37.73MiB to 171.65.103.160
20:40:19:WU01:FS01:Connecting to 171.65.103.160:8080
20:40:25:WU01:FS01:Upload 1.49%
...
20:45:24:WU01:FS01:Upload 99.05%
20:45:28:WU01:FS01:Upload complete
20:45:28:WU01:FS01:Server responded PLEASE_WAIT (464)
20:45:28:WARNING:WU01:FS01:Failed to send results, will try again later

Thanks again!

- Pete

[jdmurray]

Where do you see that? I see it is in standby (and reject) which means it's intentionally off-line. That agrees with p92xx temporarily offline, which states that the server is being worked on.

The status changed.