sunk818
|
|
March 19, 2018, 09:38:24 PM Last edit: March 22, 2018, 03:15:49 AM by sunk818 |
|
Is anyone having cgminer 4.10 cpu load go way up? If I leave it on for 12 hours or more, CPU pegs at 75%. I have 10 2Pac Gekko & 1 Terminus R808 running. Going back to Apr 2017 version of cgminer 4.10 I don't have this issue. Memory leak perhaps? At least with Apr 2017, I can run it reliably for many days. Only problem is that if Terminus goes Zombie, I have to restart it manually. Trade offs.
|
|
|
|
QuarterMaster
Newbie
Offline
Activity: 28
Merit: 2
|
Is anyone having cgminer 4.10 cpu load go way up? If I leave it on for 12 hours or more, CPU pegs at 75%. I have 10 2Pac Gekko & 1 Termins R808 running.
Going back to Apr 2017 version of cgminer 4.10 I don't have this issue.
Memory leak perhaps?
Going to guess you are running on windows? I run 4 CGMiner windows (latest build) with different groups of USB sticks/pods so I can monitor them as workers with Slushpool. About 2k memory per instance and 00-04% CPU load each, running on Server 2008R2. I do have a schedule batch file to kill all CGMiners and restart every hour though so not the best "always on" candidate here. The restart fixes the random USB stick that might have quit working. You are running the same as me minus 1 pod. I have 10 x 2pac and 2 x Terminus. Let me know if you want me to help with the batch files for killing / restarting.
|
|
|
|
|
sunk818
|
|
March 20, 2018, 12:37:23 AM |
|
Going to guess you are running on windows? Windows 10 64-bit Home. I've used shares before to quit cgminer in a batch file and loop it to restart after it quits. Say 500,000 acceptable shares then timeout for 300 seconds then start again. --shares <arg> Quit after mining N shares (default: unlimited) I have a lot of respect for what @sidehack and @vh do. It is almost magic to me what they can do in terms of hardware and driver/cgminer software programming. I'm just guessing it is some sort of memory or CPU load issue and haven't fully isolated it. Just asking if anyone here has experienced the same. If so, it is worth investigating and isolating so we can report it for a solution. The workaround of restarting is not a bad hack, but it isn't a good long-term solution.
|
|
|
|
minerbobbert
Member
Offline
Activity: 177
Merit: 10
|
|
March 20, 2018, 05:51:08 AM |
|
I bet if you only run 2 you'll get the full hashrate out of 'em. Could be a USB traffic limitation, same as several other people are running into.
Just a question I've been thinking about. Would increasing the suggest-diff setting alleviate some of the USB traffic limitations? I was just thinking, why does everyone it seems, that have posted so far, been running with really low diff? I generally run my 2pacs at 512 aiming for 2-3 Accepteds per minute, or am I doing it wrong. This is a really interesting question. For solo mining, it wouldn't even make a difference. I use ck's pools, and his pool automatically adjusts the difficulty to match the miner. Whether that impacts shares is the larger question. The USB limit is my main issue now. If this is a good workaround I'd love to know.
|
|
|
|
sidehack
Legendary
Offline
Activity: 3374
Merit: 1859
Curmudgeonly hardware guy
|
|
March 20, 2018, 12:22:49 PM |
|
I don't think suggest-diff will have any effect on traffic between cgminer and the miner itself. That only changes the threshold for returned shares to the pool.
|
|
|
|
|
sidehack
Legendary
Offline
Activity: 3374
Merit: 1859
Curmudgeonly hardware guy
|
|
March 21, 2018, 02:34:16 PM |
|
Yes, it's a hardware issue and needs warranty repair.
Speaking of which, everyone who's sent pods back for repair, they're fixed and being tested and will be return-shipped no later than Friday.
|
|
|
|
r00t$
Member
Offline
Activity: 117
Merit: 10
|
|
March 21, 2018, 03:05:46 PM |
|
Question (or request) for VH on the Gekko build...is it possible to pass in GSD for 2pac's, and GSE for pods to the --USB startup argument? I'm trying to keep multiple instances of cgminer stable and isolate the devices in bunches.
On may batch files, I'm thinking it would be something like cgminer.exe --usb GSD:4 for one instance of 4 x 2pacs, and cgminer.exe --usb GSE:4 in another instance for 4 pods. How are others separating out their Gekko devices on the same Windows OS?
|
|
|
|
goose1072
|
|
March 21, 2018, 07:36:05 PM |
|
Going to guess you are running on windows? Windows 10 64-bit Home. I've used shares before to quit cgminer in a batch file and loop it to restart after it quits. Say 500,000 acceptable shares then timeout for 300 seconds then start again. --shares <arg> Quit after mining N shares (default: unlimited) I have a lot of respect for what @sidehack and @vh do. It is almost magic to me what they can do in terms of hardware and driver/cgminer software programming. I'm just guessing it is some sort of memory or CPU load issue and haven't fully isolated it. Just asking if anyone here has experienced the same. If so, it is worth investigating and isolating so we can report it for a solution. The workaround of restarting is not a bad hack, but it isn't a good long-term solution. can you share your bat file config that does this restart when it completes say 500,000 shares? Thanks,
|
|
|
|
Namrok
Newbie
Offline
Activity: 14
Merit: 0
|
|
March 21, 2018, 08:51:51 PM |
|
So, I got my 2 pods a while ago, and they were running fun until just recently. Now they don't seem to run for more than 12 hours. It appears Rasbian itself is crapping out. I left an ssh session logged in while I was at work today to attempt to catch what happens, and this is what I saw when I came home. 6: GSD 10011790: BM1384:2 175.00MHz HW:0 | 19.01G / 18.17Gh/s WU: 253.9/m Message from syslogd@usbminer at Mar 21 09:30:52 ...---------------------------- kernel:[ 1565.872302] Internal error: Oops: 5 [#1] SMP ARMSE 0 pool 0 [2018-03-21 09:29:11.448] Accepted 207f66c5 Diff 2.02K/126 GSD 0 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...3/126 GSE 0 pool 0 kernel:[ 1565.872555] Process cg@4/Miner (pid: 1184, stack limit = 0xae0c2210) [2018-03-21 09:29:16.115] Accepted 01f8a388 Diff 130/126 GSE 1 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...2/126 GSD 0 pool 0 kernel:[ 1565.872565] Stack: (0xae0c3da8 to 0xae0c4000)126 GSE 0 pool 0 [2018-03-21 09:29:19.846] Accepted 5f04943b Diff 690/126 GSE 0 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...4/126 GSE 1 pool 0 kernel:[ 1565.872581] 3da0: b916e400 b5ed9590 a0000013 0000e2f4 b5ed95a4 b5ed9580 [2018-03-21 09:29:30.944] Accepted dad58673 Diff 299/126 GSE 1 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...0/126 GSE 1 pool 0 kernel:[ 1565.872597] 3dc0: 00000000 ae0c3ea4 b66ded00 b5ed9580 00000000 00000000 00000000 b8594900 [2018-03-21 09:29:33.838] Accepted 1c09afea Diff 2.34K/126 GSD 0 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...08K/126 GSD 2 pool 0 kernel:[ 1565.872614] 3de0: ae0c3e64 ae0c3df0 8056676c 8025dfcc 804f6744 80457afc ae0c3e14 b9a2bfc0 [2018-03-21 09:29:40.218] Accepted e1c27f9c Diff 290/126 GSD 0 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...5/126 GSE 0 pool 0 kernel:[ 1565.872630] 3e00: 00000064 ae0c3e04 ae0c3e04 ae0c3e0c ae0c3e0c b5ed9590 00000000 00000000 [2018-03-21 09:29:47.257] Accepted 01412a75 Diff 204/126 GSD 3 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...21K/126 GSE 0 pool 0 kernel:[ 1565.872645] 3e20: 00000000 00000000 ae1891c0 00000001 00000000 73600640 b92d58a4 b92d5870 [2018-03-21 09:29:53.835] Accepted 0102e5ff Diff 253/126 GSE 1 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...9/126 GSE 1 pool 0 kernel:[ 1565.872661] 3e40: 802c550a b92d58a4 b92d5800 00000015 b5ed9580 b877c9c0 ae0c3efc ae0c3e68 [2018-03-21 09:29:58.620] Accepted f203c307 Diff 271/126 GSD 4 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...6/126 GSE 0 pool 0 kernel:[ 1565.872678] 3e60: 8056710c 80565a04 8019a4d4 34406763 6e694d2f 00007265 00000000 00000000 [2018-03-21 09:30:00.932] Accepted 01c84d70 Diff 144/126 GSE 1 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...5/126 GSE 1 pool 0 kernel:[ 1565.872696] 3e80: 00000000 00000b58 73600640 00000000 00000000 00000080 6fbfc798 80108244 [2018-03-21 09:30:05.529] Accepted 01283984 Diff 221/126 GSD 2 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...1/126 GSE 0 pool 0 kernel:[ 1565.872711] 3ea0: 00000000 00008103 00000000 00000000 6fbfc8b8 00000200 00000000 00000000 [2018-03-21 09:30:23.242] Accepted 780f3291 Diff 546/126 GSE 0 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...5/126 GSE 1 pool 0 kernel:[ 1565.872726] 3ec0: 00000000 00000000 00000000 736004a0 805da2bc 73600640 b877c9c0 b5f14180 [2018-03-21 09:30:34.746] Accepted 1d56e6d2 Diff 2.23K/126 GSE 0 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...8/126 GSE 1 pool 0 kernel:[ 1565.872742] 3ee0: 802c550a 00000015 ae0c2000 00000000 ae0c3f7c ae0c3f00 802835d4 805668b0 [2018-03-21 09:30:40.033] Accepted 66dbb5eb Diff 637/126 GSE 0 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...0/126 GSD 2 pool 0 kernel:[ 1565.872758] 3f00: ae0c3f44 ae0c3f10 8018ca2c 805da53c ffffffff 00ffffff 3347ce0b 002c6192 [2018-03-21 09:30:43.469] Accepted 01e0fe6d Diff 136/126 GSE 1 pool 0 Message from syslogd@usbminer at Mar 21 09:30:52 ...8/126 GSE 0 pool 0 kernel:[ 1565.872774] 3f20: ffffffff 00000000 6fbfc788 ae0c3f88 0000004e 8028fa34 b926d000 73600668
Message from syslogd@usbminer at Mar 21 09:30:52 ... kernel:[ 1565.872792] 3f40: 73600640 00000000 802c550a 00000015 ae0c3f6c b5f14181 73600640 b5f14180
Message from syslogd@usbminer at Mar 21 09:30:52 ... kernel:[ 1565.872809] 3f60: 802c550a 00000015 ae0c2000 00000000 ae0c3fa4 ae0c3f80 80283d8c 80283534
Message from syslogd@usbminer at Mar 21 09:30:52 ... kernel:[ 1565.872825] 3f80: 01a4cde8 73600668 736004a0 00000000 00000036 80108244 00000000 ae0c3fa8
Message from syslogd@usbminer at Mar 21 09:30:52 ... kernel:[ 1565.872841] 3fa0: 801080c0 80283d54 73600668 736004a0 00000015 802c550a 73600640 01a4cde8 I've checked to make sure I have the latest drivers, and everything claims it's up to date. I'm going to try to narrow this down by going back down to stock clock, although I'm only at 175 right now, so it's not that far above stock. Also going to try one pod at a time. Anybody else seen anything like this?
|
|
|
|
sidehack
Legendary
Offline
Activity: 3374
Merit: 1859
Curmudgeonly hardware guy
|
|
March 21, 2018, 09:01:41 PM |
|
Yep, my test setups drop out a lot with the same type errors. Kinda surprised me the first time, since I had one pod running on a clean build for a couple weeks without any trouble.
|
|
|
|
|
vh (OP)
|
|
March 21, 2018, 09:51:45 PM |
|
Question (or request) for VH on the Gekko build...is it possible to pass in GSD for 2pac's, and GSE for pods to the --USB startup argument? I'm trying to keep multiple instances of cgminer stable and isolate the devices in bunches.
On may batch files, I'm thinking it would be something like cgminer.exe --usb GSD:4 for one instance of 4 x 2pacs, and cgminer.exe --usb GSE:4 in another instance for 4 pods. How are others separating out their Gekko devices on the same Windows OS?
That would be interesting, but no it's not available. In general, it would be best while testing and tuning to focus on a single unit in its own isolated environment. You'll make better progress and get more repeatable results.
|
|
|
|
Namrok
Newbie
Offline
Activity: 14
Merit: 0
|
|
March 22, 2018, 02:05:14 PM |
|
Thanks vh. So far that appears to have fixed it.
|
|
|
|
r00t$
Member
Offline
Activity: 117
Merit: 10
|
|
March 22, 2018, 06:05:03 PM |
|
Question (or request) for VH on the Gekko build...is it possible to pass in GSD for 2pac's, and GSE for pods to the --USB startup argument? I'm trying to keep multiple instances of cgminer stable and isolate the devices in bunches.
On may batch files, I'm thinking it would be something like cgminer.exe --usb GSD:4 for one instance of 4 x 2pacs, and cgminer.exe --usb GSE:4 in another instance for 4 pods. How are others separating out their Gekko devices on the same Windows OS?
That would be interesting, but no it's not available. In general, it would be best while testing and tuning to focus on a single unit in its own isolated environment. You'll make better progress and get more repeatable results. Thanks for the reply, vh. For the life of me I cannot get the 2pac's and pods to run in the same instance of cgminer, hence the question. I'm curious to know if others have been using the latest 4.10 gekko build for both sticks and pods. I can run an earlier build for just the sticks with no issues, but when I add the pods the 2pacs always zombie out.
|
|
|
|
QuarterMaster
Newbie
Offline
Activity: 28
Merit: 2
|
|
March 22, 2018, 06:43:41 PM |
|
Somebody had asked how to set up more than one Terminus on a single power supply. Here is a pic of mine running on a 200 watt adjustable supply using the PCIe connectors (very solid connection). Testing with a small fan mounted over the Regulator on the right one to keep it cool. 237.50Mhz / 98-103GH/s each. https://imgur.com/a/FPfMe
|
|
|
|
Namrok
Newbie
Offline
Activity: 14
Merit: 0
|
|
March 22, 2018, 10:57:21 PM |
|
Ok, new issue. cgminer crashed out with this error report.
Failed to create restart thread errno=12 in cgminer.c restart_threads():5063
One of my pods appears to restart a great deal, even at stock speeds. It was up to being the 229th device in cgminer. The rest were still 0 through 6.
|
|
|
|
vh (OP)
|
|
March 22, 2018, 11:54:29 PM |
|
... Failed to create restart thread errno=12 in cgminer.c restart_threads():5063 ... It was up to being the 229th device in cgminer. ...
It's unhealthy for that number to continually climb. The system will eventually prevent new resources from being allocated.. Get the other devices going in its own cgminer with command line option --usb :1, (or total devices - 1) while you work on that last one. As i mentioned a couple of post back, it would be better to test/tune it isolated.
|
|
|
|
mwin58747
Newbie
Offline
Activity: 22
Merit: 0
|
|
March 23, 2018, 02:06:13 AM |
|
... Failed to create restart thread errno=12 in cgminer.c restart_threads():5063 ... It was up to being the 229th device in cgminer. ...
It's unhealthy for that number to continually climb. The system will eventually prevent new resources from being allocated.. Get the other devices going in its own cgminer with command line option --usb :1, (or total devices - 1) while you work on that last one. As i mentioned a couple of post back, it would be better to test/tune it isolated. can this be done with linux? running multiple instances of cgminer for different devices
|
|
|
|
|