jesse11
Sr. Member
 Offline
Activity: 333
Merit: 250
Ants Rock
|
 |
January 21, 2014, 01:42:28 AM |
|
Thanks eleuthria  |
Mining with: BE's,BE Cubes, K16's, AntMiners U1's and AntMiners S1's
|
|
|
|
|
|
|
|
|
|
|
|
If you want to be a moderator, report many posts with accuracy. You will be noticed.
|
|
|
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
|
|
BrandonMcPherson
|
|
January 21, 2014, 02:27:18 AM |
|
GL with the bug finding, Eleuthria.
|
|
|
|
|
|
Trongersoll
|
|
January 21, 2014, 04:23:03 AM |
|
Yeah, my Cubes really took a hit, there running 94-97% on my end. Down 90%  I'll move them till the problem is fixed. All problems are fixed now, every BTC Guild server (including validation server) has been reverted to the same software it was using for the last year. And my machine is back to running at the proper rate. When the old stuff works and the new stuff doesn't... You have your work cut out for you. |
|
|
|
|
eleuthria (OP)
Legendary
Offline
Activity: 1750
Merit: 1007
|
|
January 21, 2014, 04:34:45 AM |
|
Yeah, my Cubes really took a hit, there running 94-97% on my end. Down 90% I'll move them till the problem is fixed. All problems are fixed now, every BTC Guild server (including validation server) has been reverted to the same software it was using for the last year. And my machine is back to running at the proper rate. When the old stuff works and the new stuff doesn't... You have your work cut out for you. It's unfortunate that the only way to find out the new stuff doesn't work is by putting it live. Every local test I can do turns out fine, but there are so many thousands of hardware configurations, software configurations, and just plain VOLUME of network activity that you can't properly test in a sterile development environment. For now, I think I might hang up this attempt at a rewrite and fall back to what I know at least one other pool has done to get better multi-core utilization: Just run the server process multiple times on multiple ports and have a proxy split the load on the same machine over the different ports. It feels like a cop out, but I noticed during the live testing that a significant load still hit one core: The one used for the primary network IO processing, which is extremely difficult to split over multiple cores/threads without significant overhead in mutex locking. I get the feeling it wouldn't be any more efficient than just running multiple instances of the server. |
RIP BTC Guild, April 2011 - June 2015
|
|
|
matt4054
Legendary
Offline
Activity: 1946
Merit: 1035
|
|
January 21, 2014, 07:02:48 AM |
|
Hey, I'm sorry for the missing log file and debugging etc. but I thought it was worth mentioning that:
ALL of my cgminer instances that were connected to BTC Guild, that is various versions from 3.0.x to the latest have crashed at the same time, approximately 06:31 GMT today 21 Jan 2014. I could see a "cannot parse output" and then a Segmentation Fault on the terminal just before crashing, while other instances just exited like they had received a SIGTERM, SIGKILL (or SIGSEGV?).
Even my Antminer S1 with factory built-in firmware running modified cgminer crashed, too. It also crashed on my KnC Miner, but this one is the only one that respawned automagically thanks to the init scripts and monitoring.
Has anyone else experienced this? I'm cross-posting this here but I also posted on the cgminer thread.
|
|
|
|
|
eleuthria (OP)
Legendary
Offline
Activity: 1750
Merit: 1007
|
|
January 21, 2014, 08:06:36 AM |
|
Hey, I'm sorry for the missing log file and debugging etc. but I thought it was worth mentioning that:
ALL of my cgminer instances that were connected to BTC Guild, that is various versions from 3.0.x to the latest have crashed at the same time, approximately 06:31 GMT today 21 Jan 2014. I could see a "cannot parse output" and then a Segmentation Fault on the terminal just before crashing, while other instances just exited like they had received a SIGTERM, SIGKILL (or SIGSEGV?).
Even my Antminer S1 with factory built-in firmware running modified cgminer crashed, too. It also crashed on my KnC Miner, but this one is the only one that respawned automagically thanks to the init scripts and monitoring.
Has anyone else experienced this? I'm cross-posting this here but I also posted on the cgminer thread.
That's...interesting. I can't think of anything that happened in the last 2 hours that could've caused such a crash. Were all the machines behind the same router? Have you been able to reconnect? |
RIP BTC Guild, April 2011 - June 2015
|
|
|
|
HellDiverUK
|
|
January 21, 2014, 09:30:22 AM |
|
cgminer is still crashing on me too. Now as soon as it connects to BTCGuild it segfaults. Same instance of cgminer has been running on Eligius perfectly for 24 hours. I've also tried it on p2pool, and also on ghash.io. It's ONLY BTCGuild that makes cgminer crash out instantly. bfgminer on the same machine connects perfectly to all pools.
NOTHING has changed in my setup, as the machine wasn't plugged in for a few days, and was working perfectly before.
|
|
|
|
|
frederikhk
Newbie
Offline
Activity: 35
Merit: 0
|
|
January 21, 2014, 09:50:13 AM |
|
can't connect with cgminer since yesterday i took a screenshot of cgminer right before it crashed: https://i.imgur.com/ibOj6fw.png |
|
|
|
|
|
HellDiverUK
|
|
January 21, 2014, 01:46:06 PM |
|
According to CK, there's a bug in cgminer which BTCGuild breaks. Yes, btcguild is a separate bug, sorry. BTCguild is the only pool that uses the redirect feature in stratum which was added blindly a long time ago when stratum support was first added and no pool used it. Cgminer's implementation is unfortunately buggy and the only workaround till I can fix it is to connect directly to btcguild's redirected url directly or use a different pool to avoid the crash.
|
|
|
|
|
eleuthria (OP)
Legendary
Offline
Activity: 1750
Merit: 1007
|
|
January 21, 2014, 05:23:41 PM |
|
Sounds like the solution for those having cgminer crashes is to downgrade to a non 3.10 version. Quite curious what the hell he did to break such a basic feature that has been working flawlessly for the last 6 months.
EDIT: You can connect to 'cgminer.btcguild.com:3333' as a temporary workaround. This server may disappear with little (or even 0) notice once the bug is known to be fixed.
|
RIP BTC Guild, April 2011 - June 2015
|
|
|
matt4054
Legendary
Offline
Activity: 1946
Merit: 1035
|
|
January 21, 2014, 05:45:54 PM |
|
Thanks for the quick server side workaround, very much appreciated. I will try it upon next crash.
I didn't have any other crash since my last post as far as I'm concerned.
|
|
|
|
|
|
AussieHash
|
|
January 21, 2014, 06:10:28 PM |
|
Sounds like the solution for those having cgminer crashes is to downgrade to a non 3.10 version. Does this affect any kncminer firmware version(s) ? |
|
|
|
|
eleuthria (OP)
Legendary
Offline
Activity: 1750
Merit: 1007
|
|
January 21, 2014, 06:12:46 PM |
|
Sounds like the solution for those having cgminer crashes is to downgrade to a non 3.10 version. Does this affect any kncminer firmware version(s) ? I have no idea. The only ASICs I have are USB Block Erupters. |
RIP BTC Guild, April 2011 - June 2015
|
|
|
matt4054
Legendary
Offline
Activity: 1946
Merit: 1035
|
|
January 21, 2014, 06:16:17 PM |
|
I'm hashing on BTC Guild with a KnC Saturn 0.99.1-tune without any issue
|
|
|
|
|
-ck
Legendary
Offline
Activity: 4088
Merit: 1631
Ruu \o/
|
|
January 21, 2014, 08:08:29 PM |
|
Sounds like the solution for those having cgminer crashes is to downgrade to a non 3.10 version. Quite curious what the hell he did to break such a basic feature that has been working flawlessly for the last 6 months.
EDIT: You can connect to 'cgminer.btcguild.com:3333' as a temporary workaround. This server may disappear with little (or even 0) notice once the bug is known to be fixed.
You didn't read the quote. It's been broken forever. |
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel
2% Fee Solo mining at solo.ckpool.org
-ck
|
|
|
eleuthria (OP)
Legendary
Offline
Activity: 1750
Merit: 1007
|
|
January 21, 2014, 08:10:20 PM |
|
Sounds like the solution for those having cgminer crashes is to downgrade to a non 3.10 version. Quite curious what the hell he did to break such a basic feature that has been working flawlessly for the last 6 months.
EDIT: You can connect to 'cgminer.btcguild.com:3333' as a temporary workaround. This server may disappear with little (or even 0) notice once the bug is known to be fixed.
You didn't read the quote. It's been broken forever. Then it must be extremely rare, because I haven't seen any reports about it causing segfaults until 3.10. BTC Guild has been using client.reconnect for a very long time. |
RIP BTC Guild, April 2011 - June 2015
|
|
|
-ck
Legendary
Offline
Activity: 4088
Merit: 1631
Ruu \o/
|
|
January 21, 2014, 08:11:48 PM |
|
Sounds like the solution for those having cgminer crashes is to downgrade to a non 3.10 version. Quite curious what the hell he did to break such a basic feature that has been working flawlessly for the last 6 months.
EDIT: You can connect to 'cgminer.btcguild.com:3333' as a temporary workaround. This server may disappear with little (or even 0) notice once the bug is known to be fixed.
You didn't read the quote. It's been broken forever. Then it must be extremely rare, because I haven't seen any reports about it causing segfaults until 3.10. The changes on it in 3.10 was my half arsed attempt to fix it when I was heading out to go overseas with little time to spare for it. |
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel
2% Fee Solo mining at solo.ckpool.org
-ck
|
|
|
eleuthria (OP)
Legendary
Offline
Activity: 1750
Merit: 1007
|
|
January 22, 2014, 05:49:06 AM |
|
There was a small bug over the last 12 hours approximately affecting the interaction between the website and backend servers when modifying worker settings like names, payment methods, and minimum difficulty. The load balancer and backends were updated to run twice as many server processes, but only the original set of processes were able to receive notifications from the website when updating worker settings. As a result, your difficulty and payment method may be different depending on which backend you connect to.
If you're experiencing this (shares paid to the wrong method or difficulty starting higher/lower than it should), you can toggle your settings back and forth to force an update to the mining servers.
This bug only affects *changes* made in the last ~12 hours. If you have not changed any settings, this would not have any affect on your miners.
|
RIP BTC Guild, April 2011 - June 2015
|
|
|
|
|
Lamalicious
Member
Offline
Activity: 148
Merit: 10
|
|
January 22, 2014, 09:15:06 AM |
|
More then 10% of the blocks. I wonder what causing it. |
|
|
|
|
