304
71%
master: 79%

Ran 30 Aug 2019 02:23AM UTC

Jobs 4

Files 69

Run time 35s

Badge

Embed ▾

pending completion

Build # 304

Build Type

push

travis-ci

Committed by

rtobar

Commit Message

Deducing rank from global list

When trying the previous solution in SUMMIT I saw that the node mapping
was not always the same through different invocations of jsrun, so the
position of a node (and therefore its IP) in the global list of IPs
given via DALIUGE_CLUSTER_IPS cannot be assumed to be the same. A
stronger check is then to let each node deduce its rank by looking the
position of its IP (i.e., my_ip) in the global list given via
DALIUGE_CLUSTER_IPS, which is common to all nodes.

With this change deployments of 6 and 200 nodes start correctly, so I
see no reason to believe that a full deployment will not work.

Signed-off-by: Rodrigo Tobar <rtobar@icrar.org>

Run Details

6925 of 9721 relevant lines covered (71.24%)

2.85 hits per line

Uncovered Existing Lines

Lines	Coverage	∆	File
2	87.53	-0.52%	dlg/utils.py

Jobs

ID	Job ID	Ran	Files	Coverage
1	304.1	30 Aug 2019 02:23AM UTC	69	71.18	Travis Job 304.1
2	304.2	30 Aug 2019 02:24AM UTC	69	71.2	Travis Job 304.2
3	304.3	30 Aug 2019 02:23AM UTC	69	71.21	Travis Job 304.3
4	304.4	30 Aug 2019 02:24AM UTC	69	71.2	Travis Job 304.4

ICRAR / daliuge / 304
71%
master: 79%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Uncovered Existing Lines

Jobs

Source Files on build 304

ICRAR / daliuge / 304 71% master: 79%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Uncovered Existing Lines

Jobs

Source Files on build 304

ICRAR / daliuge / 304
71%
master: 79%

README BADGES
x