hyouko at gmail.com
2015-Jan-06 16:13 UTC
[Nut-upsuser] Fwd: Data stale error after short while
mmh.. I looked quickly at the logs: - from time to time some character is swapped with something other (just like in the 'F' reply of your previous logs), but this shouldn't be a big problem; - then, suddenly, the queries timeout (just like when you relaunch the driver after a stale) - this is the problem. We need to make sure this isn't something vm-related (if so, having almost zero experience with it, I'm not the right person to help): - can you test it with the same os out of a vm? or - can you setup a vm with an old version of ubuntu server (I'm thinking of 10.04) and see if it behaves differently? (We had atleast one somewhat similar issue in the past - with a different driver -, claimed to be caused by usb/kernel interactions, GitHub #122) - you previously said you were not able to restart blazer_usb but successfully (albeit unintentionally) shut down the UPS after the driver stopped working: is this reproducible? Can you log the launch of blazer_usb with the 'k' switch (and still with -DDD) after the driver stops working? Also: - have you tried playing with poll interval ('i' option)? - do you see any difference if you launch, as root, the driver with the '-u root' option?
> We need to make sure this isn't something vm-relatedI suspect that this is definitely vm related now. I've tried Ubuntu server 10.10 as you suggested, and also Debian wheezy and both do the same thing. I saw in another thread that the author of the .vib of nut for ESXi is on this mailer - hopefully he can shed some light on what's happening. I would even be happy if I can simulate the cable being disconnected through the ESX CLI, have asked the same question on the vmware communities page but no answer. This is frustrating as we are about to go into a full-blown power crisis in this country and loss of data is a real risk. Thanks so far for your help, hopefully we can still get to the bottom of this though.> Also: > - do you see any difference if you launch, as root, the driver with > the '-u root' option? >I tried, and it can't re-establish the connection $ sudo /lib/nut/blazer_usb -u root -a apollo-ups -DDDDD Network UPS Tools - Megatec/Q1 protocol USB driver 0.03 (2.4.3) 0.000000 debug level is '5' 0.037486 Checking device (0665/5161) (002/004) 0.450857 - VendorID: 0665 0.450894 - ProductID: 5161 0.450900 - Manufacturer: unknown 0.450906 - Product: WayTech USB-RS232 Interface (V1.0) Baud rate 2400bps 0.450913 - Serial Number: unknown 0.450918 - Bus: 002 0.450923 Trying to match device 0.450932 Device matches 0.455814 send_to_all: SETINFO ups.vendorid "0665" 0.455852 send_to_all: SETINFO ups.productid "5161" 0.455865 send_to_all: SETINFO device.type "ups" 0.455873 send_to_all: SETINFO driver.version "2.4.3" 0.455881 send_to_all: SETINFO driver.version.internal "0.03" 0.455889 send_to_all: SETINFO driver.name "blazer_usb" 0.455895 Trying megatec protocol... 0.470826 send: Q1 1.472061 read: No error 1.472099 blazer_status: short reply 1.472107 Status read 1 failed 1.487818 send: Q1 2.490033 read: No error 2.490077 blazer_status: short reply 2.490085 Status read 2 failed 2.505990 send: Q1 3.509926 read: No error 3.509969 blazer_status: short reply 3.509977 Status read 3 failed 3.509983 Trying mustek protocol... 3.524769 send: QS 4.526100 read: No error 4.526143 blazer_status: short reply 4.526151 Status read 1 failed 4.541874 send: QS 5.544009 read: No error 5.544047 blazer_status: short reply 5.544054 Status read 2 failed 5.559816 send: QS 6.561675 read: No error 6.561714 blazer_status: short reply 6.561814 Status read 3 failed 6.561825 Trying megatec/old protocol... 6.575768 send: D 7.576204 read: No error 7.576243 blazer_status: short reply 7.576251 Status read 1 failed 7.586650 send: D 8.589290 read: No error 8.589331 blazer_status: short reply 8.589339 Status read 2 failed 8.600671 send: D 9.602361 read: No error 9.602400 blazer_status: short reply 9.602408 Status read 3 failed 9.602414 No supported UPS detected -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20150109/430e6f49/attachment.html>