Description

The articles explain the behaviour of high storage utilisation for /.mount partition when user-id process doesn't work properly because the file size of auth_v2.sdb.repl.sdb grows abnormally and this breaks the replication. As a result, the secondary node cannot sync user-id information with the primary node. Finally, the primary node gets stuck because no free storage space.

 

Symptoms

During the issue time ‘show system storage’ will fetch the below output and also the auth_v2.sdb.repl.sdb file size continue increasing.

user@root> show system storage | no-more

node0:

--------------------------------------------------------------------------

Filesystem       Size    Used   Avail Capacity  Mounted on

/dev/gpt/junos     100G    90G    2.8G    97% /.mount

tmpfs          8.3G    88K    8.3G    0% /.mount/tmp

tmpfs          2.7G    6.6M    2.6G    0% /.mount/mfs
 

user@root> file list /var/db/userid/auth_v2.sdb.repl.sdb detail

This KB also applies to the following file:
-rw-r--r--  1 root  wheel    90G Dec  9 16:30 ./var/db/userid/dpt_v2.sdb.repl.sdb

Solution

Collect the output as per the below:
 

user@root> show system storage

user@root> start shell user root

Password:

user@root% df -h

 

Check for the higher file size with the below command from shell:

user@root> start shell user root

Password:

user@root%:~ # find / -size +100000

user@root%:~ # find / -size +1000000

 

You check file size using following CLI command:

user@root> file list /var/db/userid/auth_v2.sdb.repl.sdb detail

 

In normal case, the file size will not reach 1 Gbyte. If the file size is larger than 1 Gbyte, please perform the following workaround.

Option-1:

1. Deactivate userid config using the command:
  user@root# deactivate services user-identification 

2. Run CLI command on both node0 and node1:
  user@root> file delete-directory /var/db/userid recurse

3. Activate userid config using the command:
    user@root# activate services user-identification 

 

Option-2:

1. Run CLI command on secondary node:
     user@root> file delete-directory /var/db/userid recurse

2. Perform RG0/RG1 failover

 

This issue is being tracked as PR PR1701990 .

The following software releases have been updated to resolve this specific issue:

Junos OS: 20.4R3-S8, 21.2R3-S6, 21.3R3-S5, 21.4R3-S5, 22.3R3, 22.4R2, 22.4R3, 23.1R2, 23.2R2, 23.3R1.

 

Modification History

2024-01-22 : Article Created

2025-12-18 : Can also apply to file "./var/db/userid/dpt_v2.sdb.repl.sdb"

Related Information

Related PRs: PR1701990 : https://prsearch.juniper.net/problemreport/PR1701990