Purpose of Blind Append Transactions?
388 views
Skip to first unread message
Jacek Laskowski
Apr 22, 2021, 1:51:09 PM4/22/21
to Delta Lake Users and Developers
Hi,
I've been trying to understand the purpose of so-called blind append transactions.
Are they to allow for concurrent appends only (e.g. df.format("delta").mode("append").save(...)) or are there more sophisticated use cases I'm missing? What kind of changes would not otherwise be possible?
Pozdrawiam,
Jacek Laskowski
----
Tathagata Das
Apr 22, 2021, 4:22:08 PM4/22/21
to Jacek Laskowski, Delta Lake Users and Developers
Blind appends are those unconditional data appends that do not read the contents of the table. They are marked in the commit info to identify them separately from read-modify-appends, because we do not want deletes/merges/updates to conflict with blind appends.
TD
--
You received this message because you are subscribed to the Google Groups "Delta Lake Users and Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to delta-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/delta-users/CAB_G8ZvmhSyUpP727YL3F1Yn9QM4BAzf6EG0Mwzet3xyi%2BYwxw%40mail.gmail.com.
Jacek Laskowski
Apr 23, 2021, 5:33:28 AM4/23/21
to Tathagata Das, Delta Lake Users and Developers
Hi,
> They are marked in the commit info...
...unless spark.databricks.delta.commitInfo.enabled configuration property is off, aren't they?
Is this metadata information about whether a txn is blind append or not used for concurrent transaction conflict resolution? Is this an optimization? Is this only for Serializable txns?
Pozdrawiam,
Jacek Laskowski
----
Chidananda Unchi
May 8, 2021, 10:01:15 AM5/8/21
to Delta Lake Users and Developers, apvi...@ymail.com
Reply all
Reply to author
Forward
0 new messages