Original Message:
Sent: Fri March 21, 2025 08:37 AM
From: Jan-Willem Steur
Subject: Update MAS v9.1 to v9-250306-amd64 problem
Diego,
You might try to run the updatedb.sh command shell script directly in the maxinst pod using the pod's terminal (at least that helped my to force a new update in the past)
------------------------------
Jan-Willem Steur
Manager Business Development
ZNAPZ b.v.
Breda
+31 6 25639950
Jan-Willem
Original Message:
Sent: Fri March 21, 2025 08:33 AM
From: Diego Visentin
Subject: Update MAS v9.1 to v9-250306-amd64 problem
I used mascli and I'm pretty sure the problem is not there (all steps went fine, same for pipeline).
Searching up and down, maybe I found something on the db side, so now I need to redo the updatedb task but it's not clear how to do that (documentation talks about a UPDATE_DB_CHECK variable that doesn't exist in my CRD)
------------------------------
Diego Visentin
Original Message:
Sent: Fri March 21, 2025 08:05 AM
From: Jan-Willem Steur
Subject: Update MAS v9.1 to v9-250306-amd64 problem
Diego,
I updated 3 MAS9 instances to the latest Feature Patch in 2 different OpenShift Cluster, but I didn't received this messages, neither got error in the Manage Shell app. I am using the same amd64 March version ( v9-250306) . To update to this version I used the native (docker) mascli image v13.3.0. All the Graphite (MAF) related apps are working fine without any error. The only issue I had was that MAS Assist re-install / update returned an error that the assist-secret was missing 1 critical key. I already reported this back to the Product Development team>
Did you used the ansible update approach or the mascli (docker) approach?
------------------------------
Jan-Willem Steur
Manager Business Development
ZNAPZ b.v.
Breda
+31 6 25639950
Jan-Willem
Original Message:
Sent: Thu March 20, 2025 04:01 PM
From: Diego Visentin
Subject: Update MAS v9.1 to v9-250306-amd64 problem
Hi,
when applying the latest catalog v9-250306-amd64 to a MAS v9.1 Feature Channel, I encountered a problem updating the database.
The Manage Admin Server log (i.e., the "{instance_name}-{workspace_name}-manage-maxinst-*" pod) reports several errors when updating most of the MAF applications.
For example:
MANAGE-SHELL, 9.1.0.0
// Success.
6 ms [BATCH] update mafappdata set status = 'DISABLE' where appid = ? and status = 'ACTIVE'
;
MANAGE-SHELL
// Success.
11 ms commit
;
// Success.
-1 ms [BATCH] insert into mafappdata(appid, appmode, version, buildtimestamp, deployby, deploydatetime, app, privatekey, checksum, status, revision, ismobile, mafappdataid) values (?, ?, ?, current timestamp, ?, current timestamp, ?, ?, ?, ?, ?, ?,nextval for mafappdataseq)
;
MANAGE-SHELL, DEBUG, 9.1.0.0, UPDATEDB, java.io.ByteArrayInputStream@8e6d4d35, null, ea3b9e30baccd55e1e470d4b4f0286cb, ACTIVE, 3, 0
// Failed.
rollback
This error results in no Manage app shell, effectively making it unusable.
The strange thing is that few apps turn out to be up-to-date and active, while most turn out to be the previous version disabled.
Unfortunately, I don't see the SQL code of the reported error in the log, so it's really hard for me to address get to the bottom of it.
Has anyone experienced this problem?
Any suggestions?
PS:
I know I'm testing a beta version, but this kind of problem seems too big to pass the quality check before releasing the catalog.
Or I'm making a huge mistake that I don't see ;-)
------------------------------
Diego Visentin
------------------------------