Re: migration of mysql to new server: wired errors

From: Jerry Stuckle <jstucklex_at_attglobal.net>
Date: Sun, 4 Feb 2018 19:25:39 -0600
Message-ID: <p58bos$9m1$1_at_jstuckle.eternal-september.org>


On 2/3/2018 8:46 PM, Julianony M wrote:
> On Wednesday, January 31, 2018 at 8:31:51 AM UTC-6, Jerry Stuckle wrote:

>> On 1/30/2018 11:23 PM, Julianony M wrote:
>>> On Tuesday, January 30, 2018 at 9:46:16 PM UTC-6, Jerry Stuckle wrote:
>>>> On 1/30/2018 4:55 PM, Julianony M wrote:
>>>>> I have recently migrated several databases from MySQL-5.0.95 on a RedHat-6 server to MySQL-5.6.39 on a RedHat-7 server, along with their perl applications.
>>>>>
>>>>> While everything works just fine on the old server, it encounters problems on the new server when I perform the same database load:
>>>>>
>>>>>      ERROR 1071 (42000) at line 383:
>>>>>      Specified key was too long; max key length is 767 bytes
>>>>>
>>>>> although the same data, same scripts works just fine on the old server.
>>>>>
>>>>> Could anyone suggest possible causes or spots to look at for a fix?  A simple comparisons of /etc/my.ini from the two servers didn't give me much clue.
>>>>>
>>>>> Thanks in advance!
>>>>>
>>>>> j
>>>>>
>>>>>
>>>>
>>>> How did you transfer your databases from the old to the new server?
>>>>
>>>> To start with I would show the table definitions (SHOW CREATE TABLE ...)
>>>> from the old and the new databases and compare the two.  If that doesn't
>>>> help, post both CREATE TABLE statements here along with sample data (2-3
>>>> rows should be sufficient) here.  The failing SQL statement would also
>>>> help.  We can then try to duplicate your problem.  It's hard to
>>>> troubleshoot something like this remotely.
>>>
>>> I did the database migration with 'mysqldump' into flat sql files on the old server; copy it to the new server; run mysql < sql file to set tables up.
>>>
>>> On a number of tables I compared, SHOW CREATE TABLE results from both servers are identical.  They are all "ENGINE=MyISAM DEFAULT CHARSET=latin1". (I hope this also answered Natural Philosopher's question - they are not InnodB).
>>>
>>> Thank you both for taking your time to help.
>>>
>>> j
>>>
>>
>> OK, so what does the failing CREATE TABLE statement look like?

>
>
> Sorry it took me a little while -- this is an active server with many live databases my initial capture simply included many irrelavent events. Finally I found a time window to turn others off for this test.
>
> I put the log in a Google DOC since it's very log:
> https://docs.google.com/document/d/1oWD_ksZIacNUc6Zuzawo0Op-9fUkxP_BzNGeIxrMjRY/edit?usp=sharing
> I cannot comprehend much of it. Your charp eyes and advice please? :)
>
> j
>

That wasn't the question. The question was - what does the failing CREATE TABLE statement look like? Not what was in the log file.

-- 
==================
Remove the "x" from my email address
Jerry Stuckle
jstucklex_at_attglobal.net
==================
Received on Mon Feb 05 2018 - 02:25:39 CET

Original text of this message