Changes between Version 6 and Version 7 of PS1_IPP_Czarlog_20170424

Timestamp:

Apr 30, 2017, 8:52:21 PM (9 years ago)

Author:

Mark Huber

Comment:

--

PS1_IPP_Czarlog_20170424

@@ -17 +17 @@
   * ipp001,ipp003 need mysql shutdown cleanly before halting the machines 
   * power off all machines once halted -- Gavin fixed remote power management on ippb04,b05
+  * nominal Temp to monitor -- 
+{{{
+-- ipp002 --
+Core0 Temp:
+             +44 C
+Core1 Temp:
+             +47 C
+temp1:       +41 C  (high =    +4 C, hyst =   +16 C)   sensor = thermistor   ALARM   
+temp2:     +43.5 C  (high =   +80 C, hyst =   +75 C)   sensor = thermistor           
+temp3:     +43.0 C  (high =   +80 C, hyst =   +75 C)   sensor = thermistor          
+
+-- ippops1 --
+temp1:       +36.8 C  (high = +100.0 C, hyst = +95.0 C)  sensor = Intel PECI
+temp2:      -128.0 C  (high = +100.0 C, hyst = +95.0 C)  sensor = thermal diode
+temp5:       +40.0 C  (high = +100.0 C, hyst = +95.0 C)  sensor = thermistor
+temp6:       +41.0 C  (high = +100.0 C, hyst = +95.0 C)  sensor = thermistor
+
+-- ippops2 --
+temp1:       +35.8 C  (high = +100.0 C, hyst = +95.0 C)  sensor = Intel PECI
+temp2:      -128.0 C  (high = +100.0 C, hyst = +95.0 C)  sensor = thermal diode
+temp5:       +39.0 C  (high = +100.0 C, hyst = +95.0 C)  sensor = thermistor
+temp6:       +40.0 C  (high = +100.0 C, hyst = +95.0 C)  sensor = thermistor
+}}}
+
  * MEH: ippc121 has been stalling processing @camera stage -- out of normal nightly processing for now
  * 15:45 CZW: Restarted retired host reinsertion as ipptest in screen session on stare02.  The check to confirm hosts can be retired is running slowly, as it is scanning for the existence of all files that have database entries on the retired hosts.  Reinserting will reduce the number of these that need to be checked, and the check can be resumed post-reinsertion at the point it was stopped.  This should have minimal impact, but feel free to control-C if ipp118-ipp122 loads are too high/slows down processing.
+  * the ones doing this need to monitor the level of nebulous replication slippage (looks like ~1700,3900/hr) and  adjust it accordingly if gets behind during nightly processing
+ * MEH: Haydn work at ITC
+  * ipp081 back up again -- leave neb-host down for day or so and then put into repair only until have remote power access
+  * ipp090,091 rebooted to report number of CPU properly (12 not 24) -- ipp091 raid card is dead, new ones ordered and will arrive in week or two
+  * ippdb05 needs new BBU -- wait until Monday
+  * ipp085,086,089 also need BBU but wait until db05 replaced next week
+ * MEH: ipp116.0 has gpc1 DB replication and <1TB free space.. -- should have been put into repair a while ago...
 === Saturday : YYYY.MM.DD ===