9.0.2 test conclusions

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

9.0.2 test conclusions

Phil Stracchino-2
Pretty much completed my testing.

Distinct problems found:

1.  The DB write batch size limit clearly and definitely does not work
(in addition to being set to 500000 instead of the intended 25000, per
past discussion on the subject).  This causes jobs that write more than
128K files to fail if backed by a database cluster using Galera
replication.  Since MySQL Group Replication in Oracle MySQL 5.7 is
pretty much Oracle's reverse-engineered Galera with the serial numbers
filed off, this can probably be expected to also cause problems with
MySQL Group Replication.

No problems are found when NOT using Galera, except that all attributes
are written in a single massive block at the end of the job (again
because the batching is not working).  For one of my Full backups, this
is about 4 million rows.


1a:  This should probably be a Director tunable.


2.  Storage daemon misreports available space on Solaris 11.  Available
space on an 11TB ZFS pool with 4.1TB available is reported as 1.131 PB.


3.  [Known issue]  Standalone tape drives as storage no longer work
unless configured as a fake autochanger.  How to do this correctly is
not documented.  I had to figure it out by trial and error, and what I
eventually figured out *works*, but throws errors to the console from
mtx-changer.  If the standalone tape driver issue is not readily
fixable, then at the least, the procedure for properly setting up a fake
autochanger needs to be clearly documented.



I'll file bugs on these shortly.



Carried over from prior versions:

4.  "Full backup Pool" etc. settings in jobs still override job
parameters set by the operator, but these settings are the only way to
have storage automatically selected by job level if you have different
storage for different job levels.

5.  Multiple redundant dialogs are still created by BAT when performing
certain operations.  However, it appears these can now only be responded
to in a specific order, and therefore no longer cause BAT to become
unresponsive if addressed in the wrong order.



--
  Phil Stracchino
  Babylon Communications
  [hidden email]
  [hidden email]
  Landline: +1.603.293.8485
  Mobile:   +1.603.998.6958

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Bacula-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/bacula-devel
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: 9.0.2 test conclusions

Phil Stracchino-2
On 08/01/17 15:41, Phil Stracchino wrote:
>
> I'll file bugs on these shortly.

Submitted bugs:

2299 - Database write batch size code does not work
2300 - Cannot mount or unmount tapes on standalone tape drives
2301 - Storage daemon reports available space incorrectly on Solaris 11


--
  Phil Stracchino
  Babylon Communications
  [hidden email]
  [hidden email]
  Landline: +1.603.293.8485
  Mobile:   +1.603.998.6958

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Bacula-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/bacula-devel
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: 9.0.2 test conclusions

Phil Stracchino-2
On 08/01/17 21:09, Phil Stracchino wrote:
> Submitted bugs:
>
> 2299 - Database write batch size code does not work
> 2300 - Cannot mount or unmount tapes on standalone tape drives
> 2301 - Storage daemon reports available space incorrectly on Solaris 11

Almost forgot one.

2302 - Schema not compliant with MySQL 5.7/MariaDB 10.2 because it
contains SQL errors which MySQL no longer permits.
(DATETIME fields declared NOT NULL with no DEFAULT)

DEFAULT 0000-00-00 00:00:00 will work with Bacula and MySQL 5.7/MariaDB
10.2 as long as SQL_MODE is changed to disable the default NO_ZERO_DATE
and NO_ZERO_IN_DATE restrictions.


--
  Phil Stracchino
  Babylon Communications
  [hidden email]
  [hidden email]
  Landline: +1.603.293.8485
  Mobile:   +1.603.998.6958

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Bacula-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/bacula-devel
Loading...