-
Notifications
You must be signed in to change notification settings - Fork 24
troubleshootiness 3.2 data
***********************************************************************
code: EERRUR
when: YYYY-MM-DDTHH:MM:SS.MMMZ
who: eucalyptus
what: ntpd is not running
why: cause unknown
where: the one you installed
how: can't find running ntpd process
resolution: install & run ntpd
***********************************************************************
***********************************************************************
code: EERRUR
when: YYYY-MM-DDTHH:MM:SS.MMMZ
who: account number & user id
what: instance i-%8.8X failed to run
why: libvirt is hung
where: attempt to invoke SOMEOPERATION which timed out
how: SOMEOPERATION timed out
resolution:
1. stop libvirt
2. start libvirt
3. confirm 'virsh list' returns w/o error
4. restart nc
***********************************************************************
- identifying cause: - fixing the problem: - addressing side-effects: - verification:
- ntp issues -- fault: ntpd is not running -- fault: failed to accept message due to clock skew - NC: libvirt issues when instances are being launched -- fault: running instance failed because of error returned by libvirt -- fault: running instance failed because libvirt has hung - NC: issues around volume attachment/detachment - Messaging errors in axis2c - NC: libvirt errors/NC errors in case the instance terminated prematurely - DNS setup - system ran out of some resource - traceability of resources (i.e., which CC did request go to, which NC did request go to) - quite a few errors derives from images -- having some way to keep track of a certain image boot success could be useful -- been able to check that the emi booted successfully on all the NCs -- been able to check that the emi booted successfully as all (some) users -- it booted well on all NC but X and Y -- it did booted but with a different kernel -- image/NC size limits - eucalyptus.conf sanity checks - vmware_conf sanity checks - #2177: Option to euca_conf to automatically sync eucalyptus.conf - #8726: euca_conf --initialize: check to make sure hostname is resolvable - Registration: be more verbose than "RESPONSE true" or "RESPONSE false". how about: "Registration succeeded." - host systems limitations (check and warn for disk size and provide other such hints). - image/cluster/node/hypervisor compatibility - handling of network mode/config changes - NOTREADY state error reporting - Log errors about SELinux/Apparmor related issues - error information about attempted image boot (e.g., partition table, kernel info?) - vmware image import validation - VMFS datastore configuration validation, esp. w/ multiple datastores - esxi node specific error information
- track i-123ABC - IP and EBS operations with i-123ABC - find errors associated with i-123ABC - specific handling of user-facing faults (even in log files) - per component logging - per instance (resource) logging - Installation / configuration / upgrade troubleshootiness -- generating install/config log
- #592: CONFIG: separate config files for front-end and nodes - #339: CONFIG: registration "wizard" - admin can act on any resource -- should already be the case -- when is it not? - syslogd impl.
tag:rls-3.2