MSDP configuration initiated by CloudCatalyst configuration wizard fails on dualstack media servers if the IPv6 loopback address is missing in the /etc/hosts file.
Problem
If the IPv6 loopback address is missing in the /etc/hosts file of a CloudCatalyst media server, the MSDP configuration wizard fails and rolls back the storage server configuration attempt.
Error Message
The following error message is returned when the NetBackup 8.2 Cloud Storage Configuration wizard fails:
Setting ESFS name and Local cache directory for <StorageServerName>-failed trying to exec a command(29)
Reconfiguration failed for storage server Puredisk_azure_rawd:<StorageServerName>. An error occurred in ESFS initialization
At the time of configuration, a review of the Deduplication Engine configuration log located at /<cloudcatalyst_cache>/storage/log/pdde-config.log
on the CloudCatalyst media server indicates that the Deduplication Engine (spoold) process could not be initialized within 180 seconds
Thu Nov 14 16:47:19 EST 2019 **** Done testing and init Spoold ***
Thu Nov 14 16:47:19 EST 2019 *** Direct To Cloud is disabled ***
Thu Nov 14 16:47:19 EST 2019 *** Cluster mode is disabled ***
Thu Nov 14 16:47:19 EST 2019 *** Encryption is disabled ***
Thu Nov 14 16:47:19 EST 2019 *** KMS is disabled ***
Thu Nov 14 16:47:19 EST 2019 **** Starting Content Router ***
Starting spoold
Warning: manually modifying PureDisk services can cause unexpected behavior!
Warning: press CTRL+C to abort.
Checking for PureDisk ContentRouter
spoold is stopped
spoold (pid 97032) is running...spoold [ OK ]
Thu Nov 14 16:47:31 EST 2019 Polling spoold failed. Sleeping for 10 seconds; 170 seconds remaining before timeout.
Thu Nov 14 16:47:41 EST 2019 Polling spoold failed. Sleeping for 10 seconds; 160 seconds remaining before timeout.
.....this enventually errors out......
Thu Nov 14 16:50:18 EST 2019 Polling spoold failed. Sleeping for 10 seconds; 10 seconds remaining before timeout.
Thu Nov 14 16:50:28 EST 2019 ERROR: PDDE initial configuration failed. Unable to initialize/start spoold within 180 seconds
The spoold process log located at /<cloudcatalyst_cache>/storage/log/spoold/spoold.log
on the CloudCatalyst media server may exhibit the following error messages:
November 14 16:47:21 INFO [139819644733248]: NetBindAndListen: bound myself to <localhost>:10082 using IPv6
November 14 16:47:21 INFO [139819644733248]: NetBindAndListen: Handling all connections (including IPv4, if enabled) through the IPv6 socket using IP tunneling
November 14 16:47:21 INFO [139819644733248]: Connection Manager: started
November 14 16:47:31 INFO [139817740740352]: GetSecurityProperties: synced security properties with master server 1, 0
November 14 16:47:31 ERR [139817740740352]: -1: got error response, 8618:Connection is dropped, because the host ID-to-hostname mapping is not yet approved.
November 14 16:47:31 ERR [139817740740352]: -1: validate_host_cert: validate the host ::1 uuid 0d99d3de-a234-4d69-80fa-04aee843c123 cert failed.
November 14 16:47:31 ERR [139817740740352]: 25056: Agent at ::1 certificate is invalid.
November 14 16:47:31 INFO [139817740740352]: SetCAUsageSyncFlag: set cause sync flag
November 14 16:47:31 ERR [139817740740352]: 25056: Session start request from ::1:42070 could not be honored (ssl initialization failed)
Cause
This error is seen when the default IPv6 loopback address entry in the /etc/hosts entry is missing.
::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
Solution
To resolve this error condition:
- Update the /etc/hosts file with the missing IPv6 loopback entry
- Restart the network service in Linux.
# systemctl
start|stop|restart|status
network
Please review RedHat documentation: NetworkManager and the Network Scripts for additional information.
- Once this change has been completed, cleanup the failed configuration attempt using Article: Failed to complete the initial configuration on a CloudCatalyst appliance media server
- Next use the Java GUI to re-initiate the Cloud Storage configuration Wizard