So, I joined Einstein@home, Rosetta@home and GPUgrid now, but set the resource share for Set@home to 100,000

Probably a bad idea. GPUGrid, in particular, has plenty of work at the moment - but they are long tasks with short deadlines. Their server is probably assessing - correctly - that you don't have enough time to finish their tasks at such a low relative share.

Greetings,

8696078792 	3874368987 	30 Mar 2020, 23:01:24 UTC 	31 Mar 2020, 7:34:12 UTC 	Completed, can't validate 	77.17 	75.90 	0.00 	SETI@home v8
Anonymous platform (NVIDIA GPU)
8696078794 	3875735561 	30 Mar 2020, 23:01:24 UTC 	31 Mar 2020, 7:34:12 UTC 	Validate error 	73.15 	71.74 	--- 	SETI@home v8
Anonymous platform (NVIDIA GPU)
8695980744 	3917747415 	30 Mar 2020, 22:15:09 UTC 	31 Mar 2020, 6:58:17 UTC 	Completed, can't validate 	611.40 	64.34 	0.00 	AstroPulse v7
Anonymous platform (NVIDIA GPU)
8695874883 	3875600023 	30 Mar 2020, 21:54:38 UTC 	31 Mar 2020, 5:45:30 UTC 	Completed, can't validate 	74.13 	71.96 	0.00 	SETI@home v8
Anonymous platform (NVIDIA GPU)
8695874887 	3874313675 	30 Mar 2020, 21:54:38 UTC 	31 Mar 2020, 5:45:30 UTC 	Completed, can't validate 	90.30 	88.75 	0.00 	SETI@home v8
Anonymous platform (NVIDIA GPU)
8695759325 	3875713414 	30 Mar 2020, 21:20:35 UTC 	31 Mar 2020, 4:07:51 UTC 	Completed, can't validate 	74.28 	72.02 	0.00 	SETI@home v8
Anonymous platform (NVIDIA GPU)
8695759326 	3875578801 	30 Mar 2020, 21:20:35 UTC 	31 Mar 2020, 4:07:51 UTC 	Completed, can't validate 	4.04 	1.94 	0.00 	SETI@home v8
Anonymous platform (NVIDIA GPU)

What's this "Completed, can't validate" junk? There's no such word as "can't". ;) There is some sort of 'cannot handle request' error when I try to go to the individual task page. I've not seen this type of invalidation before. LOL

Have a great day! :)

Siran

---

CAPT Siran d'Vel'nahr - L L & P _\\//
Winders 11 OS? "What a piece of junk!" - L. Skywalker
"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath

So, I joined Einstein@home, Rosetta@home and GPUgrid now, but set the resource share for Set@home to 100,000

Probably a bad idea. GPUGrid, in particular, has plenty of work at the moment - but they are long tasks with short deadlines. Their server is probably assessing - correctly - that you don't have enough time to finish their tasks at such a low relative share.

Well, since Seti@home has 0 GPU tasks availabe atm - this should play no role (I expect that there is taken into account how much time my client will actually have for computing, and not take the arbitrary resource share value with the wrong assumption that all projects always will have infinite work.). The more likely reason is that Einstein@home has completely filled up my cache for GPU tasks already.

Anyway, this was not the main point of my post. Main point was that the boinc client won't correctly use all allowed resources (empty CPU cycles for 4 cores + 50 % RAM), if there are "waiting for memory" tasks.

What's this "Completed, can't validate" junk? There's no such word as "can't". ;) There is some sort of 'cannot handle request' error when I try to go to the individual task page. I've not seen this type of invalidation before. LOL

Human error. Eric posted about this when it was first noticed: message 2041837

Anyway, this was not the main point of my post. Main point was that teh boinc client won't correctly use all allowed resources (empty CPU cycles for 4 cores + 50 % RAM), if there are "waiting for memory" tasks.

Probably a combination of several of your configuration settings. There was a bug in this area in BOINC v7.14.2, which has now been fixed. Version 7.16.5 has - at long last - been made available for official testing, and should cure it.

Anyway, this was not the main point of my post. Main point was that teh boinc client won't correctly use all allowed resources (empty CPU cycles for 4 cores + 50 % RAM), if there are "waiting for memory" tasks.

Probably a combination of several of your configuration settings. There was a bug in this area in BOINC v7.14.2, which has now been fixed. Version 7.16.5 has - at long last - been made available for official testing, and should cure it.

Seems there was really improved something in 7.16.5. It is still not using all 4 cores (only 3 atm), But if I look in taskmanager, I see that it looks like the client is really trying it's best to us the available resources. It may have some trouble with the behaviour of the Gravitational Waver Search app of using a fairly small amount of memory at the beginning and then memory usage growing up to 2 GB for one task... So it can only run 2 or 3 at the same time. ( have 6 GB , so Boinc can use 3 while I am using the computer)

What's this "Completed, can't validate" junk? There's no such word as "can't". ;) There is some sort of 'cannot handle request' error when I try to go to the individual task page. I've not seen this type of invalidation before. LOL

Human error. Eric posted about this when it was first noticed: message 2041837

Hi Richard,

If this was the fix:

I think I found the issue in a script that was supposed to trigger a resend on results unlikely to be returned. I turned the script off, so it should stop happening.

You didn't think this would go smoothly, did you?

Then why am I just now getting these reports? Did someone restart the script? :|

Have a great day! :)

Siran

---

CAPT Siran d'Vel'nahr - L L & P _\\//
Winders 11 OS? "What a piece of junk!" - L. Skywalker
"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath

Greetings,

I'm seeing something wonky in my stats pages. This result and many others are listed as "In progress". Yet, if you look at the linked page stats and click on other "In progress" results on my list, they all have the time completed and the amount of credit granted meaning they have been completed and validated. How can they be completed and validated and STILL be "In progress"? They should be listed with the valid results.

After what I see on the SSP, I believe the servers and / or database(s) are freaking out. Weird!

All of my hosts are now finished with SETI and are shut down or repurposed and my main is the only one still sitting here waiting in the hopes of getting any resends. :)

Have a great day! :)

Siran

---

CAPT Siran d'Vel'nahr - L L & P _\\//
Winders 11 OS? "What a piece of junk!" - L. Skywalker
"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath

I'm seeing something wonky in my stats pages. This result and many others are listed as "In progress". Yet, if you look at the linked page stats and click on other "In progress" results on my list, they all have the time completed and the amount of credit granted meaning they have been completed and validated. How can they be completed and validated and STILL be "In progress"? They should be listed with the valid results.

this has been discussed ad nauseam for the past month.

information in your task list come from the replica database. the replica is as of now 5 days behind (435,977 seconds). what you are seeing is the state as of 5 days ago, not "now".

---

Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

Replica seconds behind master 435,528 that's all ...


your personnal task listing is in the replica ^^ a lot of my inconclusives, waiting for wingman and in progress are done but not showed because of the replica lag ^^

---

I'm seeing something wonky in my stats pages. This result and many others are listed as "In progress". Yet, if you look at the linked page stats and click on other "In progress" results on my list, they all have the time completed and the amount of credit granted meaning they have been completed and validated. How can they be completed and validated and STILL be "In progress"? They should be listed with the valid results.

this has been discussed ad nauseam for the past month.

information in your task list come from the replica database. the replica is as of now 5 days behind (435,977 seconds). what you are seeing is the state as of 5 days ago, not "now".

Hi Ian,

Ok, that makes sense. But, I have not read anything about what I'm seeing on my stats pages in regard to this issue I mentioned, only some of what has been posted about the SSP and the replica DB. I suppose if and when the replica ever gets up to speed, we'll always see "bogus" info on stats pages. Thanks! :)

Have a great day! :)

Siran

---

CAPT Siran d'Vel'nahr - L L & P _\\//
Winders 11 OS? "What a piece of junk!" - L. Skywalker
"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath

this has been discussed ad nauseam for the past month.

Agree
What is not easy to digest is why after this almost a week with no new work the situation does not improve.
The assimilation & validation are still high, the replica time still growing, while the results in the field is about a 1/2 of last week, queries 1/3 and the reported WU are at 10K/hr only (about 1/12 of the last week).
If the load of the DB was the problem then why after remove no less than 50% of its load the same problems remains?
Does anyone has a clue? Or the real answer is a lost worst than we could imagine?.

@Ian has you still have some cache, do you have problem to report too? Mine is down now to 85k WU. Ask because i still have troubles to report the work, and makes no sense since the UL server must be practically running empty (less than 1/12 of its usual load).

---

But, I have not read anything about what I'm seeing on my stats pages in regard to this issue I mentioned

then you haven't been paying attention. all it takes is to look at the replica delay and put 2 and 2 together.

---

Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

this has been discussed ad nauseam for the past month.

Agree
What is not easy to digest is why after this almost a week with no new work the situation does not improve.
The assimilation & validation are still high, the replica time still growing, while the results in the field is about a 1/2 of last week, queries 1/3 and the reported WU are at 10K/hr only (about 1/12 of the last week).
If the load of the DB was the problem then why after remove no less than 50% of its load the same problems remains?
Does anyone has a clue? Or the real answer is a lost worst than we could imagine?.

@Ian has you still have some cache, do you have problem to report too? Mine is down now to 80k WU. Ask because i still have troubles to report the work, and makes no sense since the UL server must be practically running empty (less than 1/12 of its usual load).

the situation is slowly getting better. the assimilation numbers are slowly reducing finally. they were still over 8 million before I went to bed last night, and now down to 7.8 million. it'll just take time, and I expect that as the numbers get smaller, the speed of the recovery will increase.

I do not have any more cache, I ran out sometime on April 2nd. I've only been getting a handful of resends every day. I would just leave seti on NNT until you run out of work. I had this problem after maintenance tuesdays, I would basically have to report all of my work with NNT and let the cache go down to 0 before I got any new work.

---

Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

But, I have not read anything about what I'm seeing on my stats pages in regard to this issue I mentioned

then you haven't been paying attention. all it takes is to look at the replica delay and put 2 and 2 together.

Hi Ian,

I look at the replica delay every time I come to the website, it's the first thing I look at. As I stated, and it has NOTHING to do with "not" paying attention, I have not seen anything mentioned of what I mentioned as an "issue" in my post. I do read most of the posts about the DB and the SSP and such, some users are on my ignore list that have been for years and I don't read them. Maybe they mentioned it, I don't know.

What I mentioned as an "issue" I have not seen before and it didn't dawn on me that the "replica DB" had anything to do with it.

Have a great day! :)

Siran

---

CAPT Siran d'Vel'nahr - L L & P _\\//
Winders 11 OS? "What a piece of junk!" - L. Skywalker
"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath

Then you didn’t put 2 and 2 together. If you see that the replica is 5 days behind, and also see the dates on all the “in progress” tasks are (not) coincidentally also 5 days behind, it’s not very difficult to conclude that they might have something to do with each other. This behavior has been discussed constantly in several threads for at least the last month, and it pops up every time the replica is significantly behind.

---

Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

Looking at the tasks on my TR, I think that the replica may have reached the point that we run out of Multibeam tasks.

Hopefully the server may start to catch up again.

---

Kevin

Then you didn’t put 2 and 2 together. If you see that the replica is 5 days behind, and also see the dates on all the “in progress” tasks are (not) coincidentally also 5 days behind, it’s not very difficult to conclude that they might have something to do with each other. This behavior has been discussed constantly in several threads for at least the last month, and it pops up every time the replica is significantly behind.

I'm SORRY I friggin asked!

---

CAPT Siran d'Vel'nahr - L L & P _\\//
Winders 11 OS? "What a piece of junk!" - L. Skywalker
"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath

If the load of the DB was the problem then why after remove no less than 50% of its load the same problems remains?

50% of the load hasn't been removed.
The In progress numbers are only a very small percentage of the overall numbers, so the impact of their large reduction has been minimal.

Once we get a good large number of resends sent & returned (which will take over a month due to the deadlines), that should result in a significant drop in the "Results returned and awaiting validation" and "Workunits waiting for assimilation", and that should allow the database to finally function normally and the huge backlog of Assimilations to finally clear.

The other option is as i posted much earlier in this or another thread-
Shut down everything except for the forums, Transitioners and the MB Assimilators & let the Assimilation backlog get processed. Then start up the deleters to delete the assimilated files. Then start up the Purgers to clear them from the database.
Then startup the rest of the database, now able to function normally again, and let all those accumulated resends go out & further reduce the size of the database.

---

Grant
Darwin NT

50% of the load hasn't been removed.

Each time i asked in the past what was the way to measure the load on the DB someone tell is: queries/second

Was in the range of 1200-1500 and now is at 508 so by my math the number downs even mora than 50%

---