arma-thesis

git clone https://git.igankevich.com/arma-thesis.git
Log | Files | Refs | LICENSE

commit 3d1d1bad9e84d9cd2bde95b5e4079d789aec0055
parent 1d19e39b4cfad1a515f46bf786389e3f63d37257
Author: Ivan Gankevich <igankevich@ya.ru>
Date:   Tue, 13 Jun 2017 12:48:57 +0300

Add a table of kernel fields.

Diffstat:
arma-thesis.org | 29++++++++++++++++++++++++++++-
1 file changed, 28 insertions(+), 1 deletion(-)

diff --git a/arma-thesis.org b/arma-thesis.org @@ -3282,7 +3282,6 @@ which maximises efficiency of a programme. So, control flow objects (or kernels) possess properties of both cooperative routines and event handlers. - **** Definitions of hierarchies To disambiguate hierarchical links between daemon processes and kernels and to simplify the discussion, we will use the following naming conventions throughout @@ -3300,6 +3299,34 @@ less number of hops when kernels are distributed between cluster nodes. Since there is one-to-one correspondence between daemons and cluster nodes, they are used interchangeably in the work. +**** Kernel structure and algorithms +Each kernel has three types of fields (listed in table\nbsp{}[[tab-kernel-fields]]): +- fields related to control flow, +- fields defining the source location of the kernel, +- fields defining the current location of the kernel, and +- fields defining the target location of the kernel. + +#+name: tab-kernel-fields +#+caption: Kernel fields and their purpose. +#+attr_latex: :booktabs t :align lp{0.7\textwidth} +| Field | Description | +|-------------------+------------------------------------------------------------------------------------------------| +| ~process_id~ | Identifier of a cluster-wide process (application) a kernel belongs to. | +| ~id~ | Identifier of a kernel within a process. | +| ~pipeline~ | Identifier of a pipeline a kernel is processed on. | +| | | +| ~exit_code~ | A status of a kernel execution. | +| ~flags~ | Auxiliary flags affecting scheduling. | +| ~time_point~ | A time point at which a kernel is scheduled to be executed. | +| | | +| ~parent~ | Address/identifier of a parent kernel. | +| ~src_ip~ | IP-address of a source cluster node. | +| ~from_process_id~ | Identifier of a cluster-wide process from which a kernel was sent. | +| | | +| ~principal~ | Address/identifier of a target kernel (a kernel to which the current one is sent or returned). | +| ~dst_ip~ | IP-address of a destination cluster node. | + + ** SMP implementation **** Load balancing algorithm. The simplest approach to balance the load on a multi-processor system is to