== PS1 IPP Czar Logs for the week 2014-03-17 - 2014-03-23 == [[PageOutline]] (Up to [wiki:PS1_IPP_CzarLogs PS1 IPP Czar Logs]) === Monday : 2014-03-17 === * 09:30 Bill storage.hosts.on, storage.hosts ignore in staticsky pantasks * 11:15 at suggestion of serge and gene, kicked a diff to have a quality code of 42 === Tuesday : 2014-03-18 === mark is czar * 07:05 MEH: clearing WS diffim fault 5 (532198,532231,532259,532292 skycell.1089.062; 532381 skycell.2608.044; 533099 skycell.1084.087, skycell.1084.096) * 07:50 MEH: nightly finished, staticsky storage.hosts.on, storage.hosts.ignore * 08:20 MEH: after ~4 days, restarting regular pantasks for the week * 10:30 MEH: seeing load on ipp032 for mops and ippc30 for PSS, setting off in staticsky adds manually * ippc30 has MOPS stamps taking ~1min each rather than ~3/min normally, running staticsky on ippc30 probably isn't helping * 18:00 MEH: extra nodes for staticsky off in time for nightly, storage.hosts.off * 19:30 MEH: ipp053 looks to have gone unresponsive ~19:20.. seriously.. -- [wiki:ipp053_log] * nothing on console, power cycling since non-responsive for over 10 min, ganglia showing drop of cpu to only ~20% cpu_wait @1900 so something probably just crashed * failed to boot on first power cycle, finally kicked over on second power cycle and one of the long boot screen systems for ram check.. * nothing in logs except USB drive faults/[sr0] Unhandled sense code after reboot -- leaving in repair and out of processing for tonight === Wednesday : 2014-03-19 === mark is czar * 07:50 MEH: fault 5 diffim to clear (533929,533896 skycell.2518.043), nightly mostly done but for ~100 WS diffs * 08:40 MEH: WS diffims finished, switching nodes to staticsky except for ipp032 for Serge/MOPS * 13:30 MEH: ipp053 down again after running only staticsky load while still in neb-host repair from last night. now not booting at all on power cycle. * Haydn will look at tomorrow, setting power off * 16:00 Bill is rebuilding psModules for tests in staticsky * 17:00 MEH: rebuilding ippconfig to remove compression on the diffims for MOPS * 18:00 MEH: if extra nodes still running in staticsky, will be turning off * 19:44 Bill: fullforce processing is done (modulo faults to be investigated) setting staticsky poll limit back to 160 which is slightly larger than the number of currently enabled nodes. === Thursday : 2014-03-20 === * 09:05 Bill: storage.nodes.on, wait for all idle, storage.nodes.ignore. * rebuilt ippsky's psphot directory to fix a problem in psphotFullForce * 09:43 Bill: restarted postage stamp server pantasks. * the pstamp working directory is 96% full. This is because I turned pstamp cleanup off yesterday and so it hasn't been cleaning up after Johannes' big jobs * removed PSI and WEB labels from pstamp to give full attention to MOPS * 18:20 CZW: storage.nodes.off * 20:15 Bill: restarted summitcopy and registeration pantasks because registration was behind. ok now === Friday : 2014-03-21 === === Saturday : 2014-03-22 === === Sunday : 2014-03-23 ===