Switch to standard view 
  Sybase logo

Urgent from Sybase: Incremental backups created in a multi-writer MPX configuration using Sybase IQ 15.0, and later, may not be restorable.

Summary: This document describes a situation where Sybase IQ 15.0, and later, incremental backup archives are created without error, but the subsequent restore of these incremental backups fail to restore with an error in st_globalFlMgr::ReplayCheckpointLog ().


This document contains the following sections:

  • Customer Alert
  • Recommendation

Customer Alert

Sybase IQ 15.0 and higher can be configured in a Multiplex (MPX) configuration that contains one coordinator, multiple writers and multiple readers. This differs from previous versions that could only have a single writer and multiple readers. When Sybase IQ is configured with a single coordinator and one or more writers, incremental backups generated while there are active writers may contain incomplete checkpoint information causing restores of these incremental backups to fail.

This problem can affect any "incremental" or "incremental since full" backup in a multiplex configuration with multiple writers. Full backups and virtual backups are not affected by this issue.

This alert affects versions of Sybase IQ from 15.0 up to and including 15.1 ESD2.

This alert also affects versions of Sybase RAP - The Trading Edition from 3.0 up.


Corrective Actions for Existing Incremental Backup Archives

There is currently no way to correct incomplete incremental backups. Sybase strongly recommends that customers who use incremental backups in multi-writer MPX configurations perform a full backup or virtual backup as soon as possible.

Corrective Actions for Future Incremental Backups

After creating a full backup, any future incremental backups using affected versions of Sybase IQ should follow the steps below in order to work around for the problem.

When preparing for incremental backups, perform the following steps:

  1. Perform a normal shutdown of all writer servers. Note that killing of the writer servers does not constitute a normal shutdown. The writers servers need to be shutdown cleanly.
  2. Execute a checkpoint on the coordinator.
  3. Execute a second checkpoint on the coordinator.
  4. Run sp_iqdbspace('IQ_SYSTEM_MAIN') and verify that there are no 'G' (Global Free list Manager) type blocks in the BlkTypes column. If all writers have been shutdown correctly and there are still 'G' type blocks listed in the BlkTypes column please contact Sybase Customer Services and Support for further assistance.
    For more information on Block Types see http://infocenter.sybase.com/help/index.jsp?topic=/com.sybase.infocenter.dc38151.1510/html/iqrefbb/CIHDAEDC.htm
  5. Execute the backup command for "incremental" or "incremental since full" backup.
  6. After the incremental backup command is complete, restart the write servers.

Fixed Versions

Versions of IQ from 15.1 ESD# 3, and 15.2 , will contain a fix to correct the incremental backup issue.


Sybase is tracking this issue under CR# 625154.

This CR is fixed in the following EBFs.

Platform IQ 15.1
EBF# (ESD# 3)
IBM AIX (Power) 64-bit 17840
HP-UX (Itanium) 64-bit 17841
Linux (x86) 64-bit 17842
Linux (P-Series) 64-bit 17843
Sun Solaris (x86) 64-bit 17844
Sun Solaris (Sparc) 64-bit 17845
Windows (x86) 32-bit 17846
Windows (x86) 64-bit 17846

Customers using Sybase RAP - The Trading Edition should use the appropriate EBF for their platform from the list above


EBFs are obtained from the Sybase EBFs and Maintenance site.


Follow the instructions in the EBF cover letter to install the EBF.

If you require further assistance please contact your local support center. The contact numbers can be found in the About Support section under Support & Services at the www.sybase.com website.


Copyright © 2010 Sybase, Inc. All rights reserved.

Back to Top
© Copyright 2010, Sybase Inc.