cancel
Showing results for 
Search instead for 
Did you mean: 

Serious issue dumping a 300Gb database on ASE 15.7 Esd 4.2

Former Member
0 Kudos

Hi,

We are facing a serious issue trying to dump a 300gb database :

we tried several methods and all of them leaded the server to stop responding at all

simple dump , dump with compression=1,dump with compression=100, dump with 5 stripes.

Currently I got no way out to quiesce/rcp devices/mount the database on the target server

Sybase ASE 15.7 Esd 4.2 on Solaris 10 on Sun M3000 disks arrays SCSI on zfs

We are thinking to move the database to 2 new internal sas HD's to see if it is a problem between Sybase and Zfs.

Former Member
0 Kudos

- does the backupserver appear to be functioning properly? Absolutly

- have you ever been able to dump your database on this system, or is this your first time trying to dump the database? This system is a Main/Stby replication by log shipping working fine since last November, we perform the backups on the Stby to unload the Main server, the log shipping is performed over 3 database in a parallel way.

- have you tried bouncing the backupserver to see if this clears things up Usually we do use shutdown SYB_BACKUP, but this time we could not event connect to the dataserver

-does the dump file(s) get created? Yes and we see it growing slowly

-have you tried dumping to a non-zfs file system, even a small one? : Precisely, the devices are created on a zfs pool but we dump to a Qnap NAS (NFS) device on the lan.

-Is your ASE functioning properly, eg, are all of your databases online and accessible, can you perform read and write operations in all of your databases? Yes, since we turn the databases online on the Stby Server we encounter no problem at all.

- If you're not having any problems with the ASE itself then I'd probably postpone the movement of the dataserver to the internal HD : Well we only think that we may speed up the reads from the internal SAS HD rather than on the SCSI/ZFS, but the target device stays the NAS, according to iostat the reads are the problem, not the writes.

Thanks

Former Member
0 Kudos

Could you try dumping your master database (or an other small database) to a local HDD instead of dumping to a NFS shared directory to validate that the backup server works fine for local dumps? I've seen issues to dump databases over NFS in the past.

Accepted Solutions (0)

Answers (2)

Answers (2)

kimon_moschandreou
Contributor
0 Kudos

Hi,

this is quite interesting, the problem on the database server leading to freezing starts with

ONLINE DATABASE ems FOR STANDBY_ACCESS

For this I think someone from SAP support should help. I am afraid I cannot propose something to overcome it.

Also there is no "DUMP is complete" for emshist database. This is strange, again I cannot provide an explanation. BAckup server seems to continue after reaching the 100% of the database, however without "DUMP is complete" I am not sure if the dump file can be used to restore a database.

I am afraid I cannot help you further on this.

Regards,

Kimon

kimon_moschandreou
Contributor
0 Kudos

Hello,

Could you post the backup server log?

Best regards,

Kimon

Former Member
0 Kudos

The dump operation started at 16:35 and ended at 19:35 on database named emshist

In the bck server log you can see a weird thing : it jumps from 74% to 100% which seems abnormal and then at the same time the server dont respond anymore until I killed the process.

I could not upload the whole data server log since Im limited to 1mb per file and the file was inflated by plenty of error messages but I kept a minimum so you can have an idea.