Changes between Version 51 and Version 52 of PS1_IPP_Czarlog_20140922

Timestamp:

Sep 28, 2014, 11:09:17 PM (12 years ago)

Author:

Mark Huber

Comment:

--

PS1_IPP_Czarlog_20140922

@@ -128 +128 @@
  * 21:55 MEH: build up of nightly data but many fault 2, switching back to ipp067,068,069,070 up and ipp076,077,078,079,081,082 repair
   * should be enough space for nightly, can reboot ipp067,068,069,070 if necessary and noticed ipp078 down last night never sent email was down and so if not watching would have gone unnoticed..
-  * w/ lanl stdlocal running, tends to overload the nodes but rate seems ok still and faults not happening (wasting cycles) -- -1x c2 in lanl stdlocal seems to ease the load some (stacks?)
+  * w/ lanl stdlocal running, tends to overload the nodes but rate seems ok still and faults not happening (wasting cycles) -- -2x c2 in lanl stdlocal seems to ease the load some (stacks?)
+  * another possible alternative is to put all nodes neb-host up and use ippsXX machines to help balance against all the extra fault 2 in nightly processing, but don't want to deal with the mess if one of the no-access nodes goes down again..