From 5b53ca79c3efb3538c7c22a54d7399183d329803 Mon Sep 17 00:00:00 2001 From: Iustin Pop <iustin@google.com> Date: Fri, 8 Jul 2011 15:23:26 +0200 Subject: [PATCH] htools: update hail man page with the new modes Also mark the deprecated modes we no longer support. Signed-off-by: Iustin Pop <iustin@google.com> Reviewed-by: Michael Hanselmann <hansmi@google.com> --- man/hail.rst | 28 +++++++++++++++++----------- 1 file changed, 17 insertions(+), 11 deletions(-) diff --git a/man/hail.rst b/man/hail.rst index ca80de0d0..76380d96e 100644 --- a/man/hail.rst +++ b/man/hail.rst @@ -16,9 +16,8 @@ SYNOPSIS DESCRIPTION ----------- -hail is a Ganeti IAllocator plugin that allows automatic instance -placement and automatic instance secondary node replacement using the -same algorithm as **hbal**(1). +hail is a Ganeti IAllocator plugin that implements the instance +placement and movement using the same algorithm as **hbal**(1). The program takes input via a JSON-file containing current cluster state and the request details, and output (on stdout) a JSON-formatted @@ -30,10 +29,6 @@ ALGORITHM The program uses a simplified version of the hbal algorithm. -For relocations, we try to change the secondary node of the instance -to all the valid other nodes; the node which results in the best -cluster score is chosen. - For single-node allocations (non-mirrored instances), again we select the node which, when chosen as the primary node, gives the best score. @@ -42,11 +37,22 @@ For dual-node allocations (mirrored instances), we chose the best pair; this is the only choice where the algorithm is non-trivial with regard to cluster size. -For node evacuations (*multi-evacuate* mode), we iterate over all -instances which live as secondaries on those nodes and try to relocate -them using the single-instance relocation algorithm. +For node changes (*change-node* mode), we currently support DRBD +instances only, and all three modes (primary changes, secondary changes +and all node changes). + +For group moves (*change-group* mode), again only DRBD is supported, and +we compute the correct sequence that will result in a group change; job +failure mid-way will result in a split instance. The choice of node(s) +on the target group is based on the group score, and the choice of group +is based on the same algorithm as allocations (group with lowest score +after placement). + +The deprecated *relocate* and *multi-evacuate* modes are no longer +supported. -In all cases, the cluster scoring is identical to the hbal algorithm. +In all cases, the cluster (or group) scoring is identical to the hbal +algorithm. OPTIONS ------- -- GitLab