Lets ask a better question… why? Are you trying to clean your data? In which case, by removing the row of ‘bad’ data, you lose your auditing. (But that will depend on if you need to audit your data, and how you’re auditing it…)
Is the timestamp important? If so, why not capture it as an attribute of your data. This way you have the timestamp of the row, (hbase) and then the column. If you do this, you can capture both the timestamp associated with the data, and then the time it was entered in to your system. Does that make sense? -Mike > On Mar 23, 2015, at 9:05 AM, Ted Yu <[email protected]> wrote: > > Can you check server log to see if major compaction completed before you > issued the last put command ? > > Cheers > > On Mon, Mar 23, 2015 at 4:03 AM, mail list <[email protected] > <mailto:[email protected]>> wrote: > > > Begin forwarded message: > >> From: mail list <[email protected] <mailto:[email protected]>> >> Subject: How to put a row with an older timestamp and get it? >> Date: March 23, 2015 at 17:27:14 GMT+8 >> To: [email protected] <mailto:[email protected]> >> >> Hi ,all >> >> Recently i encounter a situation like below: >> >> >> >> That is: >> >> After i delete a row with timestamp 2015, and then put a row with timestamp >> 2014, >> How can i get the later row? >> >> > > The opinions expressed here are mine, while they may reflect a cognitive thought, that is purely accidental. Use at your own risk. Michael Segel michael_segel (AT) hotmail.com
