Import

Iam trying to import our test DS with an ldif that has around 200,000 users...The import is fine till 100,000 and afterwards , am see the following errors and the entries are getting rejected..the database is also getting corrupted....The errors are as follows:

ERROR<20775> - Backend Database - conn=9 op=120698 msgId=120780 - database error -30989

[24/Jul/2007:20:53:07 -0400] - ERROR<20501> - Backend Database - conn=-1 op=-1 msgId=-1 - Serious failure in dblayer_txn_begin, err=-30989 (D

B_RUNRECOVERY: Fatal error, run database recovery)

ERROR<33027> - Attribute Uniqueness Plugin - conn=-1 op=-1 msgId=-1 - Internal search error (1)

[24/Jul/2007:20:50:46 -0400] - ERROR<33041> - Attribute Uniqueness Plugin - conn=9 op=120585 msgId=120667 - ADD result 1

[24/Jul/2007:20:50:46 -0400] - DEBUG - conn=-1 op=-1 msgId=-1 - database index operation failed BAD 1050, err=-30989 DB_RUNRECOVERY: Fatal e

rror, run database recovery

what i absorbed is that the import is working fine upto around 60k-70k entries and afterwards am getting this problem. we are using Sun5.2sp5 on AIX...what is the exact solution for this...as am currently reinitializing the suffix as this is test environment//. urgent help is required.

[1285 byte] By [anandkaturia] at [2007-11-27 11:37:11]
# 1

Such errors during an import tend to indicate some problems with the filesystem, such as disk full, IO blocks errors or files exceeding the maximum size supported by the OS (for the system or the user).

I would look at the system's log.

Regards,

Ludovic.

ludovicpa at 2007-7-29 17:13:06 > top of Java-index,Web & Directory Servers,Directory Servers...
# 2

we have around 8GB on tht disk and currently only 20% of this space is occupied.........we are using AIX5.3..So as you said this might be a problem with IO block errors or because of the system OS..

can be you tell me what exactly is this IO blocks errors...Is it caused becaouse of the ns-slapd Process or because of the OS?

anandkaturia at 2007-7-29 17:13:06 > top of Java-index,Web & Directory Servers,Directory Servers...
# 3

Disk errors. Either OS or physical.

BTW AIX 5.3 is not a supported OS.

Regards,

Ludo

ludovicpa at 2007-7-29 17:13:06 > top of Java-index,Web & Directory Servers,Directory Servers...
# 4

thnks ludo ..but we have to go for the AIX..anyway am doing only 50k entries at each time...this way its working fine for me..

anandkaturia at 2007-7-29 17:13:06 > top of Java-index,Web & Directory Servers,Directory Servers...
# 5

because of limitations i am uploading only around 50k-60k entries at a time...but even then i got the above problem and when i tried to restart the server... i got the foll msg:

.....Server has detected a disorderly shutdown or a change in cache size

Recovery phase is starting, this may take a while...

The above thing looks and the recovery took place but when i cheked the error logs ..i get the following errors:

[27/Jul/2007:20:10:01 -0400] - DEBUG - conn=-1 op=-1 msgId=-1 - libdb: write: 0x8ae7ce48, 512: File too large

[27/Jul/2007:20:10:01 -0400] - DEBUG - conn=-1 op=-1 msgId=-1 - libdb: PANIC: File too large

[27/Jul/2007:20:10:01 -0400] - DEBUG - conn=-1 op=-1 msgId=-1 - libdb: userRoot_id2entry.db3: write failed for page 131071

What exactly is the problem for this?because We would be having around 1million+ entries and am facing these problems when i try to load with around 200k users.....We are also planning to go on live by Aug ending and so I would like to know if there are any specific things we need to take care for such a huge data import...I always get the error that the Database is getting corrupted?

Iam enabling the uniqueness plugin before Import for UID? will the import be fine i enable this plugin after Import?Please provide me a solution if anyone has come across this problem?

anandkaturia at 2007-7-29 17:13:06 > top of Java-index,Web & Directory Servers,Directory Servers...