The first factors I see from upper 900M range are from May 2008, so probably sometime around then.

I really do not know where to put this. I will put it here. If it needs moved then do so.

[QUOTE]09:32:50.271 MISFIT GPU72 work fetcher. v1.0.4 by Scott Lemieux (SWL551)
09:32:50.271
09:32:50.271
09:32:50.271 Reading program configuration from c:\gpu72\gpu72config.txt
09:32:50.271 Checking remaining work in C:\mfaktc\worktodo.txt
09:32:50.271 Found 0 assignment lines
09:32:50.271 0 is < 1: IS time to reload
09:32:50.271 Contacting GPU72 with parameters:
09:32:50.271 number=5
09:32:50.271 low=98000000
09:32:50.271 high=115000000
09:32:50.271 pledge=75
09:32:50.271 option=2
09:32:50.271 user=storm5510
09:32:50.271 pwd=xxxxxxxxx
09:32:50.271 url=https://www.gpu72.com/account/getassignments/lltf/
09:32:50.271
09:32:51.456 Logging HTML stream in GPU72FETCH_logs
09:32:51.456
09:32:51.456 Parsing HTML stream to get the assignments GPU72FETCH_logs
09:32:51.456
09:32:51.456 [B]Fatal Error: Index was out of range. Must be non-negative and less than the size of the collection.[/B]
09:32:51.456 Parameter name: [U]startIndex[/U][/QUOTE]I am having an issue with my work fetcher. "Less than the size of the collection." This seems to indicate too much distance between the [U]low[/U] and [U]high[/U] settings. If so, then where do they need to be? If not, then it is something I am not aware of.

This had been really sluggish for several days. I was only able to run about two-thirds of what I normally could because of the lost time. Manual reserve from GPU72.com is working today. Yesterday, it did not.

Anybody! :ermm:

[QUOTE=storm5510;568190]Anybody! :ermm:[/QUOTE]

Something weird is going on with the MariaDB backend. Working it.

[QUOTE=chalsall;568198]Something weird is going on with the MariaDB backend. Working it.[/QUOTE]

OK... I /think/ things are back to nominal. Had to "Optimize" the Assignment table.

Please let me know if anyone sees any further weirdness.

Colab Instance
 

Using the GPU72 TF on Colab with a CPU only session, I am looping on the following statements at varying intervals:


20210103_201124 ( 0:01): Sending Telemetry for 9c44f2a8453212b0ec80831737fb5742 (7306e513f0dae5e96200f19590c344c9)
20210103_201124 ( 0:01): Fetching payload...
20210103_201138 ( 0:02): Payload exited. Fetching next payload...

[QUOTE=linament;568220]Using the GPU72 TF on Colab with a CPU only session, I am looping on the following statements at varying intervals:


20210103_201124 ( 0:01): Sending Telemetry for 9c44f2a8453212b0ec80831737fb5742 (7306e513f0dae5e96200f19590c344c9)
20210103_201124 ( 0:01): Fetching payload...
20210103_201138 ( 0:02): Payload exited. Fetching next payload...[/QUOTE]


This is exactly true for me, too. Also CPU-only session which I just started a few minutes ago. Another CPU-only session from another account which is working on an exponent for some hours already has been giving me since one hour ago:


[CODE]20210103_215236 (11:09): [Comm thread Jan 3 21:52] Updating computer information on the server
20210103_215237 (11:09): [Comm thread Jan 3 21:52] Sending expected completion date for M105088xxx: Jan 3 2021
20210103_215237 (11:09): [Comm thread Jan 3 21:52] PnErrorResult value missing. Full response was:
20210103_215237 (11:09): [Comm thread Jan 3 21:52] [Jan 3 21:52] Visit [URL]http://mersenneforum.org[/URL] for help.
20210103_215237 (11:09): [Comm thread Jan 3 21:52] Will try contacting server again in 70 minutes.[/CODE]

Unscheduled downtime...
 

I've been fighting with this all day. I have no idea what's going on.

I'm going to take the GPU72 offline for about an hour. Sorry for the short notice, but at the moment GPU72 is effectively unusable.

Whatever change you made fixed the ranking now.

[QUOTE=ZFR;568310]Whatever change you made fixed the ranking now.[/QUOTE]

Yeah... I /thought/ perhaps the problem was the long query times caused by my Stupid DBA error earlier, so I restored the dataset to be correct.

I /think/ I've figured out the issue. An upgrade to MariaDB caused an uncommonly called Perl script to throw an error, which left a hanging lock.

[QUOTE=chalsall;568334]Yeah... I /thought/ perhaps the problem was the long query times caused by my Stupid DBA error earlier, so I restored the dataset to be correct.

I /think/ I've figured out the issue. An upgrade to MariaDB caused an uncommonly called Perl script to throw an error, which left a hanging lock.[/QUOTE]

Is everything fixed now? My assignments still show exponents that were already finished and submitted to GIMPS several hours ago.

[QUOTE=ZFR;568339]Is everything fixed now? My assignments still show exponents that were already finished and submitted to GIMPS several hours ago.[/QUOTE]

Everything /should/ be back to nominal.

Could you please PM me a couple of examples of the assignments in question?