-
Bug
-
Resolution: Duplicate
-
Critical
-
Jenkins 2.7.4, Blue Ocean 1.0.0-b07
-
-
pacific
MN Summary: more information is needed than is exposed currently as steps in blue ocean (and possibly classic, but that isn't clear just yet) - more investigation needed. An unpleasant fallback is to resort to viewing full log, but that isn't obvious or nice.
It appears that the console output doesn't include the "node allocation" information which is useful when trying to debug systemic issues in Jenkins agents.
Take this Jenkinsfile which can execute on one of a number of agents. We had some issues where an agent was misbehaving, resulting in failures like:
WARNING: Failed to execute: select name,version,instanceid from plugin where month >= 1466636926735 and name NOT LIKE 'privateplugin%' and version NOT LIKE '%(private)' order by name,version desc,instanceid because: [SQLITE_IOERR] Some kind of disk I/O error occurred (disk I/O error)
LOLWUT aside, the beginning of the "classic" Console Output view includes:
Started by user rtyler
Replayed #8
Connecting to https://api.github.com using jenkinsadmin/****** (GitHub access token for jenkinsadmin)
[Pipeline] node
Running on kelp in /home/jenkins/agent-workspace/workspace/Infra/infra-statistics/master
[Pipeline] {
[Pipeline] checkout
Fetching changes from 2 remote Git repositories
Checking out Revision 4c2c3e9a24627421e7cd17d725a7a42bfe5a06e7 (master)
[Pipeline] tool
[Pipeline] tool
[Pipeline] stage (Sync raw data and census files)
A screenshot from the Blue Ocean detail page for this run is attached. Suffice it to say, there is no information in the detail page about what nodes this Pipeline run executed on 
- duplicates
-
JENKINS-40466 Developer wants to see a message when Jenkins is busy eg: execution is blocked waiting for a node
-
- Resolved
-
- relates to
-
-
- Closed
-
-
-
- Closed
-
- links to
[JENKINS-38759] Results screen/nodes in Blue Ocean omits important information
| Attachment | New: Screen Shot 2016-10-06 at 7.04.24 pm.png [ 34269 ] |
| Description |
Original:
It appears that the console output doesn't include the "node allocation" information which is useful when trying to debug systemic issues in Jenkins agents. Take [this Jenkinsfile|https://github.com/jenkins-infra/infra-statistics/blob/master/Jenkinsfile] which can execute on one of a number of agents. We had some issues where an agent was misbehaving, resulting in failures like: {code} WARNING: Failed to execute: select name,version,instanceid from plugin where month >= 1466636926735 and name NOT LIKE 'privateplugin%' and version NOT LIKE '%(private)' order by name,version desc,instanceid because: [SQLITE_IOERR] Some kind of disk I/O error occurred (disk I/O error) {code} LOLWUT aside, the beginning of the "classic" Console Output view includes: {code} Started by user rtyler Replayed #8 Connecting to https://api.github.com using jenkinsadmin/****** (GitHub access token for jenkinsadmin) [Pipeline] node Running on kelp in /home/jenkins/agent-workspace/workspace/Infra/infra-statistics/master [Pipeline] { [Pipeline] checkout Fetching changes from 2 remote Git repositories Checking out Revision 4c2c3e9a24627421e7cd17d725a7a42bfe5a06e7 (master) [Pipeline] tool [Pipeline] tool [Pipeline] stage (Sync raw data and census files) {code} A screenshot from the Blue Ocean detail page for this run is attached. Suffice it to say, there is no information in the detail page about what nodes this Pipeline run executed on :( |
New:
MN Summary: more information is needed than is exposed currently as steps in blue ocean (and possibly classic, but that isn't clear just yet) - more investigation needed. An unpleasant fallback is to resort to viewing full log, but that isn't obvious or nice. It appears that the console output doesn't include the "node allocation" information which is useful when trying to debug systemic issues in Jenkins agents. Take [this Jenkinsfile|https://github.com/jenkins-infra/infra-statistics/blob/master/Jenkinsfile] which can execute on one of a number of agents. We had some issues where an agent was misbehaving, resulting in failures like: {code} WARNING: Failed to execute: select name,version,instanceid from plugin where month >= 1466636926735 and name NOT LIKE 'privateplugin%' and version NOT LIKE '%(private)' order by name,version desc,instanceid because: [SQLITE_IOERR] Some kind of disk I/O error occurred (disk I/O error) {code} LOLWUT aside, the beginning of the "classic" Console Output view includes: {code} Started by user rtyler Replayed #8 Connecting to https://api.github.com using jenkinsadmin/****** (GitHub access token for jenkinsadmin) [Pipeline] node Running on kelp in /home/jenkins/agent-workspace/workspace/Infra/infra-statistics/master [Pipeline] { [Pipeline] checkout Fetching changes from 2 remote Git repositories Checking out Revision 4c2c3e9a24627421e7cd17d725a7a42bfe5a06e7 (master) [Pipeline] tool [Pipeline] tool [Pipeline] stage (Sync raw data and census files) {code} A screenshot from the Blue Ocean detail page for this run is attached. Suffice it to say, there is no information in the detail page about what nodes this Pipeline run executed on :( |
| Assignee | New: Vivek Pandey [ vivek ] |
| Epic Link | New: JENKINS-35759 [ 171771 ] |
vivek I thought this may be of interest while you are digging into bismuth. It would be wonderful if we could surface this extra information as steps. If this isn't possible, perhaps we should close this and either rethink the design to show the full log in some cases, or ... don't know what else.
| Sprint | New: atlantic [ 106 ] |
| Sprint | Original: atlantic [ 106 ] | New: pacific [ 101 ] |
| Priority | Original: Minor [ 4 ] | New: Major [ 3 ] |
201604291_tyler it would be good if you could also look at the classic stage view - to see if it includes it there (I think it includes some detail)
Based on primitive experiments, it looks like it may include more.