GPU72/Colab Error
 

I don't know whether this belongs in cloud computing or GPU72.

Anyways, I've been running colab fine until I got no TF assignments for the GPU, I was connected to host runtime. I restarted got an error and still it's not getting work. Any ideas as to whats happening?

FileNotFoundError
 

I might be hijacking your thread, but currently GPU72/Colab is giving me a FileNotFoundError. Perhaps this is related to your issue? Screen shot is attached. Thanks.

[QUOTE=Runtime Error;554084]I might be hijacking your thread, but currently GPU72/Colab is giving me a FileNotFoundError. Perhaps this is related to your issue? Screen shot is attached. Thanks.[/QUOTE]

I think it it we have the same gzip and tar error, or whatever that would be called

GPU72 site certificate
 

There is a problem with the GPU72 site certificate. I get an error when trying to access the site.

I got the same thing on two different accounts.

[QUOTE]Beginning GPU Trial Factoring Environment Bootstrapping...
Please see [url]https://www.gpu72.com/[/url] for additional details.


gzip: stdin: unexpected end of file
tar: Child returned status 1
tar: Error is not recoverable: exiting now
---------------------------------------------------------------------------
FileNotFoundError Traceback (most recent call last)
<ipython-input-1-f14dc9c16595> in <module>()
24 stdout=subprocess.PIPE,
25 universal_newlines=True,
---> 26 bufsize=0)
27
28 try:

1 frames
/usr/lib/python3.6/subprocess.py in _execute_child(self, args, executable, preexec_fn, close_fds, pass_fds, cwd, env, startupinfo, creationflags, shell, p2cread, p2cwrite, c2pread, c2pwrite, errread, errwrite, restore_signals, start_new_session)
1362 if errno_num == errno.ENOENT:
1363 err_msg += ': ' + repr(err_filename)
-> 1364 raise child_exception_type(errno_num, err_msg, err_filename)
1365 raise child_exception_type(err_msg)
1366

FileNotFoundError: [Errno 2] No such file or directory: './bootstrap.pl': './bootstrap.pl'[/QUOTE]

Argh, my throughput! /wrists

[QUOTE=]20200818_023839 INFO: Comms spider starting...
20200818_023839 INFO: Gracefull shutdown...
Use of uninitialized value $AID in concatenation (.) or string at ./comms.pl line 309.
20200818_023839 ERR : Bad response:
500 Can't connect to [url]www.gpu72.com:443[/url] (certificate verify failed)
[/QUOTE]

Yep, it's a certificate problem

Same multiple self-tests here. Same error message.
I am relieved that it is a systemic problem. I had been signing in to more than one account, and trying to set up Colab in a new gmail account. When things started acting crazy, I thought I had screwed something up, and I ran through the routine of saving and shutting down all the notebook tabs, signing out of all the accounts, and back into my main account, and calling up the notebooks again. Alas, to no avail.

Oh well. I still need to clean up the kitchen. Something to pass the time

[QUOTE=Awesomeotts;554092]Yep, it's a certificate problem[/QUOTE]


Same thing happened to me the other day on one of my websites. Those "Let's Encrypt" certificates that most hosting providers are using freely for their shared hosts are "supposed" to auto renew.


I had to initiate an online chat with my provider (HostGator) and the guy installed a new cert on the server on the spot, and everything went back to normal.

Cert updated...
 

Sorry guys... I stupidly let the cert expire... Updated.

Without your commitment, we won't be able to earn those Colab Ghz/day !
So no reason to be sorry ...

:goodposting:

Something clogging up the works this evening? Lots of 100% assignments hanging around having a party.

[QUOTE=LOBES;558823]Something clogging up the works this evening? Lots of 100% assignments hanging around having a party.[/QUOTE]
That started with me sometime in the last day, 10-30-2020. Right now, my Overall Statistics says I have 405 TF assignments. The P-1 numbers are in the normal range: teens. The TF numbers are still going up. :help:
EDIT: "Overall" is back to the usual range. View Assignments is still high, but it always runs behind Overall.

Not related to any of the above, but there's now a warning at the top:[quote][color=orangered]output-height is not a valid annotation. Supported annotations are: vertical-output, run, form-width, display-mode[/color][/quote]

[QUOTE=James Heinrich;566048]Not related to any of the above, but there's now a warning at the top:[/QUOTE]

Yeah... They're messing with us now...

Rather than silently ignoring settings they used to accept, but don't any longer, they're now presenting these warnings.

Since I'm "radiating", I haven't seen a GPU across any of my 13# Colab accounts since last Wednesday. And even the CPU-only instances last for only about 30 minutes each.

Thanks for all the Factors Google. But, yeah, Colab is effectively dead-in-the-water for GPU72 efforts. :sad:

[QUOTE=chalsall;566052]Yeah... They're messing with us now...
...Colab is effectively dead-in-the-water for GPU72 efforts. :sad:[/QUOTE]Strange. I have multiple non-gpu72 free Colab sessions going, and some just passed 4 hours of gpu now. A couple others ceased around the 6 hour mark. But I've noticed day of the week seems to matter.

[QUOTE=chalsall;566052]Since I'm "radiating", I haven't seen a GPU across any of my 13# Colab accounts since last Wednesday. And even the CPU-only instances last for only about 30 minutes each.[/QUOTE]There is words to the effect of "lifetime usage limits". This is what I think is happening with regards to GPUs. On the account attached to my phone I got a T4 for about 45 minutes eariler today. It has far less hours on it than the other account that I have used.

[QUOTE=Uncwilly;566109]There is words to the effect of "lifetime usage limits". This is what I think is happening with regards to GPUs.[/QUOTE]

Yeah, but the resolution in the equation granting access is finer than simply each Human has the same Lifetime Usage Limits (LULs (LOL...)).

Just about every other GPU72_Colab Notebook user started having serious constraints well before I personally did. And, somewhat ironically, it was the Paid Tier who got shut-down first.

I think it also has something to do with the Notebook being run, as there are reports of others continuing to have access to Colab GPU access using different code-paths.

And, last data-set... A certain Greg continues to have access to four (4#) GPU72_Colab instances consistently, and others intermittently.

Google, like ${DEITY}, acts in strange ways. Almost quantum (or, at least, intractable)...

[QUOTE=LOBES;558823]Something clogging up the works this evening? Lots of 100% assignments hanging around having a party.[/QUOTE]


Ditto here. I FINALLY get a GPU and this happens. :huh: :mad: Additionally, at least as of right now, when getting a GPU it ONLY does TF. Not TF and P-1 like it usually does.