Page 1 of 1

SQLite DB error executing BEGIN: TRANSACTION;~INSERT INTO

Posted: Mon Nov 20, 2023 7:06 am
by celoxgroup
Hi all,

System: 3PAR 8400 3.3.2.159 (MU1)+P10

I don't know when exactly this problem happened but it might be after I have applied the latest updates.

Second controller node is reporting SQLite DB error executing ... My suspicion is that the Insight database is corrupted on the node controller 1. I could not find instructions how to reinstate the database and hopefully someone knows how to do it. No DB errors on node 0.

I get a few errors every 5 minutes :/ and looks like this:

Code: Select all

SQLite DB error executing BEGIN: TRANSACTION;~INSERT INTO current.vvspace VALUES (1700480700,'-',0,0,'admin','60002AC000000000000000000001F878','---','---','full','base','','RW','normal','2023-07-15 19:02:40 AEST','--',10240,0,0,0,0,0,0,0,0,0,0,10240,20480,10240,0,0,20971520.0,20971520.0,0,0,0,0,'-','-','-',0,'-',0,0,'NA',0,1700480700,0);~INSERT INTO current.vvspace VALUES (1700480700,'-',1,1,'.srdata','60002AC000000000000000010001F878','---','---','full','base','','RW','normal','2023-07-15 19:03:00 AEST','--',112640,0,0,0,0,0,0,0,0,0,0,11.

Recommended Action
Description: An internal database used to track and take action based on system statistics is not responding as expected. Resolution: Contact your next level of support.

Component
Detected by node 1


Kind Regards, Matt

Re: SQLite DB error executing BEGIN: TRANSACTION;~INSERT INTO

Posted: Mon Sep 30, 2024 5:52 am
by celoxgroup
Hi all,

Still struggling with the error mentioned, however at least I know where this error is coming from :)

Seems like that the internal SQLite database for the Node 1 is corrupted on the System Reporter (.srdata) virtual volume. I wonder if anyone has experience how to recreate/reformat that volume. I don't mind if I lose the historic metrics.

I came across of this article (https://3parug.com/viewtopic.php?t=2749) and tried to stop the service, admithw srdata however I got the same error message as the user "System Reporter data volume exists." which makes sense as it does exist and the volume itself is in a healthy state.

Google-fu is failing me and seems like no one has this problem that I could find of.

Sudden observation as showsr outputs only Node 1. Is this correct? Or should there be also Node 0?

Code: Select all

3par cli% showsr
Node Total(MiB) Used(MiB) Used%
-------------------------------
   1     118937    104443    93
I have shortened the retention period just now:

Code: Select all

3par cli% showsr
Node Total(MiB) Used(MiB) Used%
-------------------------------
   1     118937    104443    93

Filetype info:
                 -(MiB)- -Target- --------Retention---------
FileType   Count   Usage   Period Target Max Estimate Actual EarliestDate        EndEstimate
----------------------------------------------------------------------------------------------------
aomoves        1       0      ---    --- ---      ---    --- ---                 ---
baddb          0       0      ---    --- ---      ---    --- ---                 ---
daily          1     745       1d     1d ---     10y+  1.21y 2023-07-16 00:00:00 10+ years from now
hires         86   85859       5m     1d ---     353d   244d 2024-01-31 06:35:00 9.4 days from now
hourly        18   17161       1h     1d ---     353d  1.21y 2023-07-15 20:00:00 Full or almost full
ldrg           0       0 disabled    --- ---      ---    --- ---                 ---
perfsample     4       3      ---    --- ---      ---    --- ---                 ---
srmain         1       0      ---    --- ---      ---    --- ---                 ---
system         0       0      ---    --- ---      ---    --- ---                 ---
unused       ---    8452      ---    --- ---      ---    --- ---                 ---
Disabled data types: ldrg
Thank you and kind regards, Matt