Skip to content

TS-4929: No loading of HostDB disk file if sync_frequency=0#1076

Merged
zwoop merged 1 commit into
apache:masterfrom
zwoop:TS-4929
Oct 6, 2016
Merged

TS-4929: No loading of HostDB disk file if sync_frequency=0#1076
zwoop merged 1 commit into
apache:masterfrom
zwoop:TS-4929

Conversation

@zwoop
Copy link
Copy Markdown
Contributor

@zwoop zwoop commented Oct 4, 2016

This has two benefit (2) is most important I think:

  1. We avoid warnings on startup about not being able to load the HostDB.

  2. Since in 7.0.0, turning off syncing (sync_frequency=0) completely disables writing to HostDB, it's now possible to end up loading a potentially stale HostDB from a previous upgrade.

@zwoop zwoop added the HostDB label Oct 4, 2016
@zwoop zwoop added this to the 7.1.0 milestone Oct 4, 2016
@atsci
Copy link
Copy Markdown

atsci commented Oct 4, 2016

FreeBSD build successful! See https://ci.trafficserver.apache.org/job/Github-FreeBSD/928/ for details.

@atsci
Copy link
Copy Markdown

atsci commented Oct 4, 2016

Linux build successful! See https://ci.trafficserver.apache.org/job/Github-Linux/822/ for details.

Copy link
Copy Markdown
Contributor

@PSUdaemon PSUdaemon left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This looks good to me. BTW, adding a ?w=1 to the diff page makes it much simpler to review.

@jpeach
Copy link
Copy Markdown
Contributor

jpeach commented Oct 5, 2016

@zwoop (1) and (2) sound like they are in conflict? I'm confused about whether it loads or not? If it does load, do you really have a stale record problem, since each record has a TTL?

@zwoop
Copy link
Copy Markdown
Contributor Author

zwoop commented Oct 5, 2016

Yes, they would have TTLs, but if I turned off syncing, I certainly wouldn't expect it to load something old. What if I had bad entries? What if I had a corruption of some sort? Since it will never update it, you are stuck with loading old (possibly stale or more likely expired) entries for the remaining of life? That just makes no sense.

@jpeach jpeach assigned jpeach and unassigned jpeach Oct 5, 2016
@jpeach
Copy link
Copy Markdown
Contributor

jpeach commented Oct 5, 2016

OK that sounds reasonable.

@zwoop zwoop merged commit 16a4067 into apache:master Oct 6, 2016
@zwoop zwoop deleted the TS-4929 branch October 6, 2016 16:18
bneradt added a commit to bneradt/trafficserver that referenced this pull request Jan 23, 2026
* unique_ptr<char[]> -> array; default initialize .value

* SnowflakeIDGenerator

This encapsulates the state for snowflake id generation in a generator,
avoiding global mutable state.

---------

Co-authored-by: Brian Neradt <brian.neradt@gmail.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants