This message was deleted.
# opals
Slackbot
07/17/2023, 1:19 AMThis message was deleted.
o
Or Weis
07/17/2023, 5:38 AMHi @Oscar OPAL doesn't persist the data, but instead retchs it when the client restarts.
I'm guessing what you are experiencing is a race condition between the policy update(which can contain data in data.json, and the regular data updates
Or Weis
07/17/2023, 5:42 AMSee this recent conversation: https://permit-io.slack.com/archives/C01RUUYV3TP/p1689263308953139
o
Oscar
07/17/2023, 5:43 AMI mean if there are some data updates by 'opal-client publish-data-update', then we will lost all those updates once restart the server.
Oscar
07/17/2023, 5:44 AMThen we have to keep those updates and do them again after server running?
Oscar
07/17/2023, 5:44 AMOtherwise, we only can update data by git repo.
o
Or Weis
07/17/2023, 5:44 AMYes, but the client will refetch according to
DATA_SOURCES_CONFIGo
Oscar
07/17/2023, 5:45 AMyes, it needs to keep full data in data sources.
o
Or Weis
07/17/2023, 5:48 AMAnother option is to use offline mode (store backup and reuse) with
OPAL_OFFLINE_MODE_ENABLED=1o
Oscar
07/17/2023, 5:49 AMOk, let me take a look. Thanks.
🤘 1
✅ 1
r
Ro'e Katz
07/17/2023, 3:23 PM@Or Weis @Oscar
Using the offline mode for that is a cool idea, but use it with caution.
The main purpose of this feature is to avoid downtime if client goes up and server is not reachable for any reason. Make sure that new data (from singular push updates) won’t be overridden by older data after restart once connection to server is established (if your server’s git repo / data sources points to older data on the same paths).
✅ 1
o
Oscar
07/18/2023, 1:11 AM@Ro'e Katz thanks
p
permit support automation2
07/18/2023, 1:11 AMHi, thanks for reaching out 😇;
It seems most of the team is AFK(😴) at the moment, so our response might be delayed.
2 Views