How far back will publisher syncronize?
Purged all data for a synced data source. and asked the ES publisher to sync all 216 points history back to April 1st. 2017. The publisher reported this..
10:49:29: Points: 216, sent 216 requests, synced 12072213 records, target overcounts: 0, response errors: 0, elapsed time: 1m33s
How can 1m33s be correct? and the target dataset remains empty and has not changed.
Is it possible the receiver isn't configured to create points (Accept point updates)? Do you have anything useful in the logs?
To the question in the title, the Persistent TCP publisher tracks points individually for how far their history synchronization has gotten. There is a tool on the publisher page to set a start time for all data points on the publisher. The sync would be between the time set there and the current moment minus the "synchronize prior to" setting.
Yes I am familiar with the publisher settings as it's been used successfully for a couple months now and normally keeps the system up to date and fills in the histories nightly etc. However since I screwed up the new noSQL database, I put the mysql db back to the cloud while I work on the noSQL db on my desktop. so one of our ES is piping the current data to the cloud on one pipe and on a second pipe I am asking it to sync the entire history to my empty desktop database with noSQL running on that server to get a fresh noSQL database. Then I will move that to the cloud server. sound right?
[removing for having a public IP]
Yeah, that sounds like it should work fine.
my prior to is 2 hours
[removing because public IP]
When the publisher status reports ... History sync not currently running
Does that mean that it might still run again or has it finished the sync?
It means it is not running. The synchronize cron pattern will determine when it runs next, and it looks like yours is empty, so it will not run unless manually started.
You can enable logging on the data source and the publisher, as well, if we need more info.
Yes I realize I start it manually but does that mean it may stop before it has completed?
Sorry about the ips I forget sometimes.. Even though the publisher reports successful syncs no records are in the target dataset since I purged it SO now I am going to delete the datasource completely and recreate it using the same identifiers.
We shall see if rebuilding the dataset from scratch corrects the situation.
I believe I had to do this one time before to get the history working properly on a pipe..
Finally got the ES syncing all history to noSQL on cloud now by deleting and re-creating the data source and using diff on the mango folder I found the deleted files and restored them.
All the ES data is back on the cloud now in enterprise noSQL db and ES pipes are up and running. Meta calculations also have been running all night and about 1/3 through the dataset this morning so all in all a good result.
Glad to hear it!