http://mining.bitcoin.cz/accounts/profile/ - 502 bad gateway again

[Veldy]

Slush - is your bitcoind daemon down again?  What is your plan to fix this issue?

Thank you.

[rezin777]

There is an entire thread dedicated to this pool.

http://bitcointalk.org/index.php?topic=1976.msg97363#msg97363

[Veldy]

There is an entire thread dedicated to this pool.

http://bitcointalk.org/index.php?topic=1976.msg97363#msg97363

There sure is and I have been part of it.  This is more specific however.  This is a recurring problem and a specific problem and not general pool discussion.  Some threads should be allowed to die and not resurrected forever because the subject is related to such a general previous topic.

Slush and Tycho really should setup their own forums for their pools, but then again, if Slush host it himself, it would be down right now ....

[Veldy]

BTW, as of the time of this post, the site is back up and running.

[LMGTFY]

Slush and Tycho really should setup their own forums for their pools, but then again, if Slush host it himself, it would be down right now ....

I dunno. I quite like having discussion here in the open, in a central location. I use more than pool (I use BTCmine as well as Slush's) and neither have their own forum. It always struck me as a bit odd that bitcoinpool closed and locked the pool thread here and moved to their own forum. It kind of creates a chicken-and-egg situation: I'm not aware of issues with their pool, so I don't use their pool. I don't use their pool so I don't follow their forum. I don't follow their forum so I'm not aware of issues with their pool... etc

Back on topic, as you note the 502 issue is a recurring one, but it doesn't bother me. It doesn't affect my mining, all it affects is my stats recording and that recovers as soon as the 502 issue is resolved. It just strikes me as a low priority issue. Or am I missing something?

[Veldy]

Slush and Tycho really should setup their own forums for their pools, but then again, if Slush host it himself, it would be down right now ....

I dunno. I quite like having discussion here in the open, in a central location. I use more than pool (I use BTCmine as well as Slush's) and neither have their own forum. It always struck me as a bit odd that bitcoinpool closed and locked the pool thread here and moved to their own forum. It kind of creates a chicken-and-egg situation: I'm not aware of issues with their pool, so I don't use their pool. I don't use their pool so I don't follow their forum. I don't follow their forum so I'm not aware of issues with their pool... etc

Back on topic, as you note the 502 issue is a recurring one, but it doesn't bother me. It doesn't affect my mining, all it affects is my stats recording and that recovers as soon as the 502 issue is resolved. It just strikes me as a low priority issue. Or am I missing something?

I am referring mostly to maintenance, outage and service issues and other things that users of the pool are asking about.  There is nothing stopping discussion of any of the pools here.  Sometimes specifics don't need to be posted here if their was a forum at each pool.

[slush]

Please always discuss pool troubles on pool thread - I'm busy and often don't read rest of the forum for many days.

I'm finishing other projects with strict deadlines those days, I'm going to sleep at 5am every day, so for now I'm restarting bitcoind manually every time the hangs happen (it was twice yesteday and once today, but it usually keeps running many days without any problem). It is easier than spending one or two evenings on patching & testing & deploying new version of bitcoind. Those hangs are ugly, but they don't break mining.

[Veldy]

Please always discuss pool troubles on pool thread - I'm busy and often don't read rest of the forum for many days.

I'm finishing other projects with strict deadlines those days, I'm going to sleep at 5am every day, so for now I'm restarting bitcoind manually every time the hangs happen (it was twice yesteday and once today, but it usually keeps running many days without any problem). It is easier than spending one or two evenings on patching & testing & deploying new version of bitcoind. Those hangs are ugly, but they don't break mining.

Take a look at monitoring software to detect the hang and to restart the daemon; when you have the time. 

You are correct, mining went on normally.

Thanks.

[LMGTFY]

Slush and Tycho really should setup their own forums for their pools, but then again, if Slush host it himself, it would be down right now ....

I dunno. I quite like having discussion here in the open, in a central location. I use more than pool (I use BTCmine as well as Slush's) and neither have their own forum. It always struck me as a bit odd that bitcoinpool closed and locked the pool thread here and moved to their own forum. It kind of creates a chicken-and-egg situation: I'm not aware of issues with their pool, so I don't use their pool. I don't use their pool so I don't follow their forum. I don't follow their forum so I'm not aware of issues with their pool... etc

Back on topic, as you note the 502 issue is a recurring one, but it doesn't bother me. It doesn't affect my mining, all it affects is my stats recording and that recovers as soon as the 502 issue is resolved. It just strikes me as a low priority issue. Or am I missing something?

I am referring mostly to maintenance, outage and service issues and other things that users of the pool are asking about.  There is nothing stopping discussion of any of the pools here.  Sometimes specifics don't need to be posted here if their was a forum at each pool.

I guess I just don't see the point of either (a) having discussion elsewhere (here works), or (b) attaching a high priority to what seems to me to be an unimportant issue. Help me understand why this matters to you, is all I'm saying.

[Veldy]

I guess I just don't see the point of either (a) having discussion elsewhere (here works), or (b) attaching a high priority to what seems to me to be an unimportant issue. Help me understand why this matters to you, is all I'm saying.

It keeps day to day operational issues out of this forum.  They can keep it here if they want; I just think it is better if bundled with their service.  One gigantic thread that never ends is simply STUPID.

MY OPINION (no, I am not yelling, just stressing the issue).

[LMGTFY]

I guess I just don't see the point of either (a) having discussion elsewhere (here works), or (b) attaching a high priority to what seems to me to be an unimportant issue. Help me understand why this matters to you, is all I'm saying.

It keeps day to day operational issues out of this forum.  They can keep it here if they want; I just think it is better if bundled with their service.  One gigantic thread that never ends is simply STUPID.

MY OPINION (no, I am not yelling, just stressing the issue).

OK, well I agree about the "one gigantic thread", though I think it makes sense to continue discussions in one thread (this topic seems to be a continuation of an earlier discussion). What I don't get is (minor issue) why have the discussion elsewhere, and (major issue) why is the 502 issue that important? It doesn't affect mining, we know that, we've known that for a while. It just doesn't seem an important issue to me, and I'm still struggling to see why it is important for you.

[slush]

Well, I understand why are people asking - it really looks weird, I know. But unfortunately it is minor issue which need a lot of time, which I'm currently don't have. I hope people understand this, I'm also fixing that ASAP.

[Veldy]

I guess I just don't see the point of either (a) having discussion elsewhere (here works), or (b) attaching a high priority to what seems to me to be an unimportant issue. Help me understand why this matters to you, is all I'm saying.

It keeps day to day operational issues out of this forum.  They can keep it here if they want; I just think it is better if bundled with their service.  One gigantic thread that never ends is simply STUPID.

MY OPINION (no, I am not yelling, just stressing the issue).

OK, well I agree about the "one gigantic thread", though I think it makes sense to continue discussions in one thread (this topic seems to be a continuation of an earlier discussion). What I don't get is (minor issue) why have the discussion elsewhere, and (major issue) why is the 502 issue that important? It doesn't affect mining, we know that, we've known that for a while. It just doesn't seem an important issue to me, and I'm still struggling to see why it is important for you.

It affects payout, access to your account and frankly accountability [which I am not too worried about, but there is a lot of hacking going around].  Any one of these reasons is enough and I just listed three.

[LMGTFY]

OK, well I agree about the "one gigantic thread", though I think it makes sense to continue discussions in one thread (this topic seems to be a continuation of an earlier discussion). What I don't get is (minor issue) why have the discussion elsewhere, and (major issue) why is the 502 issue that important? It doesn't affect mining, we know that, we've known that for a while. It just doesn't seem an important issue to me, and I'm still struggling to see why it is important for you.

It affects payout, access to your account and frankly accountability [which I am not too worried about, but there is a lot of hacking going around].  Any one of these reasons is enough and I just listed three.

Well, that's your opinion (that one of these reasons is enough). A delayed payout doesn't concern me. Access to my account is the statistics issue I mentioned earlier and, frankly, that's so trivial as to be risible. Accountability? C'mon. Wait until the server comes back up and check if you're worried. Better yet, have your miners set up so they switch pools when you get a 502 from Slush's, and let the remaining pool participants worry about accountability.

Molehill -> mountain.

[Veldy]

It affects payout, access to your account and frankly accountability [which I am not too worried about, but there is a lot of hacking going around].  Any one of these reasons is enough and I just listed three.

Well, that's your opinion (that one of these reasons is enough). A delayed payout doesn't concern me. Access to my account is the statistics issue I mentioned earlier and, frankly, that's so trivial as to be risible. Accountability? C'mon. Wait until the server comes back up and check if you're worried. Better yet, have your miners set up so they switch pools when you get a 502 from Slush's, and let the remaining pool participants worry about accountability.

Molehill -> mountain.

Well, that is YOUR opinion.  If I find the need to verify or change my address for any reason, I can't.  There is a delay for payout while blocks are confirmed.  If I just switch pools, I may not get paid if I can't login and set the payout threshold low enough to trigger a payment when my work is confirmed.  Bitcoins, while a recent invention, is money and there is a duty to make this safe and secure and accessible.  Granted, pooling is all new and if you keep the payout threshold low and keep a close eye on things, the risk of loss is also low.

As far as your mountain out of a molehill comment; that is off base as I didn't say anything about it being a major problem.  At the moment, it is annoying (and it is already back up).  However, when hackers get in [as they did with, was it BitcoinPool?], then it is a big deal if you don't keep close watch.  This becomes real money for people and as such, the pools should be maintained with that in mind.

Slush promised to work on it and I am happy with that for now.  I just reported the outage.  He should have monitoring software to report it to him on its own so that users don't have to report, but perhaps that is part of his upcoming solution.  The only other item I mentioned is a dedicated forum for each pool.  I suggested on their sites, but it could be here if they want to create that many .. or perhaps a few for the bigger pools and an "Other" for the smaller pools.

There was no building of mountains here.  Slush is the pool I started with and the pool I am currently using.  That will not change for awhile now [unless there is a pool outage].  It seems my work is somehow disappearing in small amounts (seemingly network related, but doesn't explain the share count differences between my client and the statistics onsite) when I use deepbit, so I am doing a long test with slush for now since I am overdue for the long term stats anyway.  Also, I like the idea of scoring (modified proportional really).

There is an awful lot of trust involved with pooled mining and even more so when server security and stability is at stake.  So, I care.  Sorry if you think that is a mountain.

[Veldy]

It's down again.  

[LMGTFY]

Well, that is YOUR opinion.  If I find the need to verify or change my address for any reason, I can't.  There is a delay for payout while blocks are confirmed.  If I just switch pools, I may not get paid if I can't login and set the payout threshold low enough to trigger a payment when my work is confirmed.  Bitcoins, while a recent invention, is money and there is a duty to make this safe and secure and accessible.  Granted, pooling is all new and if you keep the payout threshold low and keep a close eye on things, the risk of loss is also low.

Fair enough. I take the view that much more of the responsibility for this lies with us than with Slush (or other pool operators). We need to make sure that our mining operations are safe, and rely less on pool operators. For example (and this doesn't apply to you, or, indeed, to Slush) one operator is promising to notify participants when one of their miners goes down. I think this is... not a mistake, but is perhaps less helpful than it sounds. What I think is needed is for participants to monitor their miners locally, rather than relying on pool operators. Operators should provide services like this with large disclaimers and warnings that pool participants have responsibility for monitoring their own operations.

As far as your mountain out of a molehill comment; that is off base as I didn't say anything about it being a major problem.  At the moment, it is annoying (and it is already back up).  However, when hackers get in [as they did with, was it BitcoinPool?], then it is a big deal if you don't keep close watch.  This becomes real money for people and as such, the pools should be maintained with that in mind.

Slush promised to work on it and I am happy with that for now.  I just reported the outage.  He should have monitoring software to report it to him on its own so that users don't have to report, but perhaps that is part of his upcoming solution.  The only other item I mentioned is a dedicated forum for each pool.  I suggested on their sites, but it could be here if they want to create that many .. or perhaps a few for the bigger pools and an "Other" for the smaller pools.

There was no building of mountains here.  Slush is the pool I started with and the pool I am currently using.  That will not change for awhile now [unless there is a pool outage].  It seems my work is somehow disappearing in small amounts (seemingly network related, but doesn't explain the share count differences between my client and the statistics onsite) when I use deepbit, so I am doing a long test with slush for now since I am overdue for the long term stats anyway.  Also, I like the idea of scoring (modified proportional really).

There is an awful lot of trust involved with pooled mining and even more so when server security and stability is at stake.  So, I care.  Sorry if you think that is a mountain.

Well, apologies if you object to the "mountains" comment. I made it because you started a new thread to discuss an issue that was already being discussed elsewhere. It wasn't clear to me what, if any, information Slush could give you that you weren't already aware of. That was why I tried to find out what, exactly, concerned you.

I think we agree on more than we disagree on, we just have different approaches to the issues we face.

[slush]

Veldy, I wrote simple watchdog monitoring bitcoind and restarting it when it hangs. Hope it will solve situation till Monday, then I'll compile patched bitcoind.

[Veldy]

Veldy, I wrote simple watchdog monitoring bitcoind and restarting it when it hangs. Hope it will solve situation till Monday, then I'll compile patched bitcoind.

Thank you.  I think that will make a lot of pool members feel better; especially if they want coins sent regularly.  I noticed that the new bitcoind fixes the known lockup issue that you have been experiencing.  Have you considered suggesting your patches to the bitcoind project for acceptance into the main code base, or are they proprietary to you pool?

[slush]

I noticed that the new bitcoind fixes the known lockup issue that you have been experiencing.

We talk about upgrading to the newest bitcoind all the time .

Have you considered suggesting your patches to the bitcoind project for acceptance into the main code base, or are they proprietary to you pool?

Well, they are custom patches like "don't close rpc connection after every request" or "report current block number in getwork response" and so on, which is not needed for other deployments outside the pool. There are also some custom patches from other people (like jgarzik's settxfee or m0mchil's getwork lockup solution), which are publicly available, but are not in main source tree (yet).