mersenneforum.org  

Go Back   mersenneforum.org > Factoring Projects > FactorDB

Reply
 
Thread Tools
Old 2023-02-25, 15:22   #606
Happy5214
 
Happy5214's Avatar
 
"Alexander"
Nov 2008
The Alamo City

98010 Posts
Default

When uploading an aliquot sequence elf file using aliqueit, I can only upload 70-80 terms at a time. This is a new issue, since the switch to partial solar power (though I haven't done aliquot work in several months, so I don't know if it started earlier). This is quite an annoyance, since my sequence uploads are normally automated by my working environment, and this is making me double back and re-upload the sequences in several blocks.
Happy5214 is offline   Reply With Quote
Old 2023-03-01, 11:32   #607
gd_barnes
 
gd_barnes's Avatar
 
"Gary"
May 2007
Overland Park, KS

22×17×179 Posts
Default

Quote:
Originally Posted by Happy5214 View Post
When uploading an aliquot sequence elf file using aliqueit, I can only upload 70-80 terms at a time. This is a new issue, since the switch to partial solar power (though I haven't done aliquot work in several months, so I don't know if it started earlier). This is quite an annoyance, since my sequence uploads are normally automated by my working environment, and this is making me double back and re-upload the sequences in several blocks.
This was an update ~2-3 weeks ago so that people could not upload tens of thousands of ID's all at once to circumvent the hourly limits. You can only upload ~600 IDs at a time. I just manually add them 70-80 indexes at a time now. You can go over by a few hundred ID's by adding them up to 5900+ ID's in an hour and then dropping 70-80 indexes, which amount to 500-600 ID's, and then wait for the next hour.

The DB is much better now. With far less spamming of pointless 10^n+c and 10^n-c forms, response is better and we don't have to wait hours sometimes for the brown IDs to go away for proofs because the DB is bogged down by such spam. A very much needed change.
gd_barnes is offline   Reply With Quote
Old 2023-03-01, 16:23   #608
EdH
 
EdH's Avatar
 
"Ed Hall"
Dec 2009
Adirondack Mtns

3·7·263 Posts
Default

For automated longer sequence runs I've added a separate script that watches the .elf and uploads for each additional term added.
EdH is offline   Reply With Quote
Old 2023-03-01, 17:23   #609
Happy5214
 
Happy5214's Avatar
 
"Alexander"
Nov 2008
The Alamo City

22·5·72 Posts
Default

Quote:
Originally Posted by gd_barnes View Post
This was an update ~2-3 weeks ago so that people could not upload tens of thousands of ID's all at once to circumvent the hourly limits. You can only upload ~600 IDs at a time. I just manually add them 70-80 indexes at a time now. You can go over by a few hundred ID's by adding them up to 5900+ ID's in an hour and then dropping 70-80 indexes, which amount to 500-600 ID's, and then wait for the next hour.

The DB is much better now. With far less spamming of pointless 10^n+c and 10^n-c forms, response is better and we don't have to wait hours sometimes for the brown IDs to go away for proofs because the DB is bogged down by such spam. A very much needed change.
At long last! I'll gladly take the added inconvenience with uploading aliquot sequences if it means the spamming is curtailed. I just wish that had been announced (unless I just missed it).
Happy5214 is offline   Reply With Quote
Old 2023-03-01, 17:26   #610
Tyler Busby
 
Tyler Busby's Avatar
 
Jan 2023

61 Posts
Default

Quote:
Originally Posted by gd_barnes View Post
This was an update ~2-3 weeks ago so that people could not upload tens of thousands of ID's all at once to circumvent the hourly limits. You can only upload ~600 IDs at a time. I just manually add them 70-80 indexes at a time now. You can go over by a few hundred ID's by adding them up to 5900+ ID's in an hour and then dropping 70-80 indexes, which amount to 500-600 ID's, and then wait for the next hour.

The DB is much better now. With far less spamming of pointless 10^n+c and 10^n-c forms, response is better and we don't have to wait hours sometimes for the brown IDs to go away for proofs because the DB is bogged down by such spam. A very much needed change.
This must be why the smallest composite can be up to 77 digits easily now. Previously the number would continuously be much lower due to "spam".

Speaking of, the distribution graph shows 2-4 unfactored composites of size 13, 57, (and maybe 74) that aren't returned by the "unfactored composite" API. I assume these are "phantom" composites caused by either a caching or data corruption issue, and their existence makes contributing computing power to factoring composites marginally less satisfying (which is, of course, chief motivation for the undergoing).

I do wonder now what smallest composite size can realistically be achieved with the new limits.

I had a script running t20 ecm against 79-digit composites yesterday, to attempt to smooth the corner there. But I seemed to only net a single pixel at that point in the scale (which, I estimate to be around 290000 composites, with 80 being ~280000). And the API only lets you pick from the first 50000, so at a certain point of this strategy I may have raised the average roughness of the selection pool by a significant amount. I think my script averaged about one successful factorization every 2 seconds, so that's roughly 27 hours of running before it's seen 50000.

Also, to bring this post back on topic, it seems there are problems with the server now, just as I'm finishing writing.

Code:
Proxy Error
The proxy server received an invalid response from an upstream server.
The proxy server could not handle the request

Reason: Error reading from remote server
Likely not purely on my side as it's happening to my colab sessions too.

Last fiddled with by Tyler Busby on 2023-03-01 at 17:29
Tyler Busby is offline   Reply With Quote
Old 2023-03-01, 18:17   #611
EdH
 
EdH's Avatar
 
"Ed Hall"
Dec 2009
Adirondack Mtns

3×7×263 Posts
Default

Quote:
Originally Posted by Tyler Busby View Post
. . .
Also, to bring this post back on topic, it seems there are problems with the server now, just as I'm finishing writing.

Code:
Proxy Error
The proxy server received an invalid response from an upstream server.
The proxy server could not handle the request

Reason: Error reading from remote server
Likely not purely on my side as it's happening to my colab sessions too.
I'm seeing this, too.
EdH is offline   Reply With Quote
Old 2023-03-01, 19:32   #612
Stargate38
 
Stargate38's Avatar
 
"Daniel Jackson"
May 2011
14285714285714285714

769 Posts
Default

This is what came up when the DB finally responded. DB outages shouldn't count toward the quota. Why do they even count as CPU time?! This is what I was trying to load: http://www.factordb.com/index.php?id...00004378897973. I'm factoring the C134 via SNFS atm.

I was hitting F5 every 10-20 minutes to see when the DB finally responded, but that shouldn't have ramped up the CPU to nearly 2 hours like that. IOW, when an HTTP 5xx error (server issue) occurs, the time needed for the error to appear shouldn't be counted towards anyone's quota, regardless of what number they were loading.
Attached Thumbnails
Click image for larger version

Name:	Screenshot 2023-03-01 142854.png
Views:	28
Size:	19.2 KB
ID:	28080  

Last fiddled with by Stargate38 on 2023-03-01 at 19:39
Stargate38 is offline   Reply With Quote
Old 2023-03-01, 20:18   #613
Tyler Busby
 
Tyler Busby's Avatar
 
Jan 2023

3D16 Posts
Default

For what it's worth, I think clicking on the status page starts some infinitely hanging process that contributes to one's parallel processing quota, as well as maybe doing a search. Perhaps the db is in a bad state, but the "limits" page seems okay.
Tyler Busby is offline   Reply With Quote
Old 2023-03-03, 05:15   #614
gd_barnes
 
gd_barnes's Avatar
 
"Gary"
May 2007
Overland Park, KS

101111100011002 Posts
Default

Quote:
Originally Posted by Happy5214 View Post
At long last! I'll gladly take the added inconvenience with uploading aliquot sequences if it means the spamming is curtailed. I just wish that had been announced (unless I just missed it).
I didn't see a formal announcement. There was a temporary message on the site that stated something to the effect of: "Due to spam, the number of IDs that can be submitted at any one time has been reduced to 500.". (I'm not sure if that was the limit but it was something close to that.) It gave the impression that the change would be temporary but a day or two later, the message was gone and it was clear that the change had become permanent.

Since I sometimes have multiple machines running aliquot sequences from 100 to 120 digits 24x7, after running overnight I'm often way over the limit that I can submit in an hour. I just have to space out the updates over several sessions during the day.

It's a small price to pay to prevent all of the terrible spam. The DB now only has the rare hiccup, as would be expected for such a large CPU-intensive operation, but for the most part it performs excellently now.
gd_barnes is offline   Reply With Quote
Old 2023-03-10, 03:44   #615
Happy5214
 
Happy5214's Avatar
 
"Alexander"
Nov 2008
The Alamo City

22·5·72 Posts
Default

Is it just me, or is the spam making a comeback?

Last fiddled with by Happy5214 on 2023-03-10 at 03:44 Reason: Use actual icon
Happy5214 is offline   Reply With Quote
Old 2023-03-10, 14:37   #616
richs
 
richs's Avatar
 
"Rich"
Aug 2002
Benicia, California

6A816 Posts
Default

The spammer is back. Over 500,000 composites of the form 10^k-n added in the past two days....
richs is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
A suggestion for factordb. enzocreti FactorDB 24 2022-11-17 07:20
Extending Factordb carpetpool FactorDB 6 2017-01-23 11:04
FactorDB PRP's smh FactorDB 231 2015-07-28 02:30
bugged sequence in factordb firejuggler Aliquot Sequences 2 2010-06-15 14:03
FactorDB question Raman Factoring 15 2010-01-28 10:24

All times are UTC. The time now is 05:04.


Sun Jun 4 05:04:13 UTC 2023 up 290 days, 2:32, 0 users, load averages: 0.88, 1.06, 1.08

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2023, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.

≠ ± ∓ ÷ × · − √ ‰ ⊗ ⊕ ⊖ ⊘ ⊙ ≤ ≥ ≦ ≧ ≨ ≩ ≺ ≻ ≼ ≽ ⊏ ⊐ ⊑ ⊒ ² ³ °
∠ ∟ ° ≅ ~ ‖ ⟂ ⫛
≡ ≜ ≈ ∝ ∞ ≪ ≫ ⌊⌋ ⌈⌉ ∘ ∏ ∐ ∑ ∧ ∨ ∩ ∪ ⨀ ⊕ ⊗ 𝖕 𝖖 𝖗 ⊲ ⊳
∅ ∖ ∁ ↦ ↣ ∩ ∪ ⊆ ⊂ ⊄ ⊊ ⊇ ⊃ ⊅ ⊋ ⊖ ∈ ∉ ∋ ∌ ℕ ℤ ℚ ℝ ℂ ℵ ℶ ℷ ℸ 𝓟
¬ ∨ ∧ ⊕ → ← ⇒ ⇐ ⇔ ∀ ∃ ∄ ∴ ∵ ⊤ ⊥ ⊢ ⊨ ⫤ ⊣ … ⋯ ⋮ ⋰ ⋱
∫ ∬ ∭ ∮ ∯ ∰ ∇ ∆ δ ∂ ℱ ℒ ℓ
𝛢𝛼 𝛣𝛽 𝛤𝛾 𝛥𝛿 𝛦𝜀𝜖 𝛧𝜁 𝛨𝜂 𝛩𝜃𝜗 𝛪𝜄 𝛫𝜅 𝛬𝜆 𝛭𝜇 𝛮𝜈 𝛯𝜉 𝛰𝜊 𝛱𝜋 𝛲𝜌 𝛴𝜎𝜍 𝛵𝜏 𝛶𝜐 𝛷𝜙𝜑 𝛸𝜒 𝛹𝜓 𝛺𝜔