This is observed in HADR where " During the takeover, the virtual IP and HA proxy are moved to the new node and then the DB2 HADR takeover is performed, Meaning the DB2 is on from the Standbyserver and backup attempt during the time will fail with such error.
Standby databases can be backedup and configured via netvault plugin to avoid backup failure during standby ON State
IBM reference to check the current state of HADR database server
when run on the primary, the cmwo_db2hadr_status command will show HADR_ROLE=PRIMARY for all databases, HADR_STATE=PEER for the first HADR standby (also known as the principal standby), and HADR_STATE=REMOTE_CATCHUP for all other HADR standbys. When run on a standby node, the HADR_ROLE will be STANDBY, and the HADR_STATE will be PEER or REMOTE_CATCHUP. The cmwo_db2hadr_status command will display summary output similar to the following output:
*** Summary: ***
All databases are PRIMARY
DB2 HADR looks OKAY.
https://www.ibm.com/support/knowledgecenter/en/SST55W_4.3.0/liaca/liaca_operation_not_allowed_onhadr_db.html