log backup: checkpoint cannot advance after pause->stop->start

[AkiraXie]

Bug Report

Please answer these questions before submitting your issue. Thanks!

1. Minimal reproduce step (Required)

1. start a log backup
2. pause it
3. stop it
4. start a new one

2. What did you expect to see? (Required)

normal

3. What did you see instead (Required)

checkpoint cannot advance, but status shows normal

              name: longrun-pitr-test
            status: ● NORMAL
             start: 2024-05-06 02:20:40.574 +0000
               end: 2090-11-18 14:07:45.624 +0000
           storage: s3://qe-testing/kernel-testing/longrun/log-0506
       speed(est.): 0.00 ops/s
checkpoint[global]: 2024-05-06 02:20:40.574 +0000; gap=5h25m28s

4. What is your TiDB version? (Required)

v8.1

[AkiraXie]

/component br
/severity critical

[AkiraXie]

/label affects-8.1