Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

kvs: log date of restored checkpoint #3580

Closed
garlick opened this issue Mar 31, 2021 · 3 comments
Closed

kvs: log date of restored checkpoint #3580

garlick opened this issue Mar 31, 2021 · 3 comments
Assignees

Comments

@garlick
Copy link
Member

garlick commented Mar 31, 2021

It would be handy if this log message at startup

kvs.info[0]: restored kvs-primary from checkpoint

included the date the checkpoint was written.

Possibly related to #2783

@grondo
Copy link
Contributor

grondo commented Mar 31, 2021

Also, related to #3552?

@garlick
Copy link
Member Author

garlick commented Mar 31, 2021

Oops that's the issue I meant to reference, not that merged pr. Duh.

@chu11 chu11 self-assigned this Feb 15, 2022
chu11 added a commit to chu11/flux-core that referenced this issue Feb 15, 2022
Problem: It'd be convenient if we knew the date when the kvs
primary checkpoint was checkpointed.

Solution: When checkpointing the primary KVS, store a json
object with both the rootref and timestamp, instead of just
the rootref string.  On retrieval, parse appropriately and
retrieve timestamp for output in logs.

Fixes flux-framework#3580
chu11 added a commit to chu11/flux-core that referenced this issue Feb 15, 2022
Problem: It'd be convenient if we knew the date when the kvs
primary checkpoint was checkpointed.

Solution: When checkpointing the primary KVS, store a json
object with both the rootref and timestamp, instead of just
the rootref string.  On retrieval, parse appropriately and
retrieve timestamp for output in logs.

Fixes flux-framework#3580
chu11 added a commit to chu11/flux-core that referenced this issue Feb 16, 2022
Problem: It'd be convenient if we knew the date when the kvs
primary checkpoint was checkpointed.

Solution: When checkpointing the primary KVS, store a json
object with version, rootref, and timestamp, instead of just
the rootref string.  On retrieval, parse appropriately and
retrieve timestamp for output in logs.  Support the original
checkpointing format by checking if the checkpoint object
is a raw blobref string first.

Fixes flux-framework#3580
chu11 added a commit to chu11/flux-core that referenced this issue Feb 17, 2022
Problem: It'd be convenient if we knew the date when the
primary namespace was checkpointed.

Solution: When checkpointing the primary namespace, store a json
object with version, rootref, and timestamp, instead of just
the rootref string.  On retrieval, parse appropriately and
retrieve timestamp for output in logs.  Support the original
checkpointing format by checking if the checkpoint object
is a raw blobref string first.

Fixes flux-framework#3580
chu11 added a commit to chu11/flux-core that referenced this issue Feb 17, 2022
Problem: It'd be convenient if we knew the date when the
primary namespace was checkpointed.

Solution: When checkpointing the primary namespace, store a json
object with version, rootref, and timestamp, instead of just
the rootref string.  On retrieval, parse appropriately and
retrieve timestamp for output in logs.  Support the original
checkpointing format by checking if the checkpoint object
is a raw blobref string first.

Fixes flux-framework#3580
chu11 added a commit to chu11/flux-core that referenced this issue Feb 17, 2022
Problem: It'd be convenient if we knew the date when the
primary namespace was checkpointed.

Solution: When checkpointing the primary namespace, store a json
object with version, rootref, and timestamp, instead of just
the rootref string.  On retrieval, parse appropriately and
retrieve timestamp for output in logs.  Support the original
checkpointing format by checking if the checkpoint object
is a raw blobref string first.

Fixes flux-framework#3580
@mergify mergify bot closed this as completed in e1b0450 Feb 18, 2022
@garlick
Copy link
Member Author

garlick commented Feb 18, 2022

Just wanted to note here that I installed flux-core containing #4136 on my test system instance and restarted, and it successfully parsed the old format checkpoint, logging

2022-02-18T23:40:53.124149Z kvs.info[0]: restored kvs-primary from checkpoint on N/A

Then I ran systemctl restart flux and noted:

2022-02-18T23:43:03.383331Z kvs.info[0]: restored kvs-primary from checkpoint on 2022-02-18T23:43:03

So that's working as designed. Good job @chu11.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants