RME Database Inconsitency -

getwithrob · ‎03-26-2006

Is there an easy, non-intrusive way to see if the RME database is loaded with inconsistencies?

The Solaris box is running RME 3.5 only with no other CW apps.

The reason I ask is the box has been acting kind of weird lately. This morning I had to do a reboot because it showed no devices when I went into Hardware Inventory (I’ve actually seen this before but a reboot always corrected the problem).

I did a reboot then waited several minutes and tried to perform a manual device add and saw an error on the screen I had never seen before. I then did a self-test and saw where database.pl had failed. I rebooted again and came back to it right at an hour later and tried to add the device again. After I put the device name and other attributes in, it told me the DIServer process was down when it got to the import status screen. I was able to go to server config > Administration and manually start the DIServer process that showed down using pdshow. After this, everything seems to be OK so I thought I would post to see if there’s a way to check the DB for inconsistencies while the server is running.

If I've ever seen problems before, a reboot would always correct the problem. I

Note: This RME box has over 4,500 devices in the inventory.

getwithrob · ‎03-26-2006

I found this in a previous post regarding the same topic from nhabib. I have not tried this and am a bit reluctant to since the DB is so very large.

Has anyone tried this?

If so how long did it take to run? How many devices were in the RME database at the time?

*****************************************************

Here are the instructions to correct inconsistencies. Of course, make sure you perform a backup before running them.

You need to be root and using a Bourne or Korn shell

/etc/init.d/dmgtd stop

LD_LIBRARY_PATH=/opt/CSCOpx/lib:/opt/CSCOpx/objects/db/lib

export LD_LIBRARY_PATH

PATH=${PATH}:/opt/CSCOpx/objects/db/bin ; export PATH

SATMP=/tmp/.SQLAnywhere ; export SATMP

ASANY=/opt/CSCOpx/objects/db ; export ASANY

cd /opt/CSCOpx/databases/rme

rm rme.log (if it exists)

dbsrv7 -f rme.db

/etc/init.d/dmgtd start

****************************************************

yjdabear · ‎03-27-2006

As far as the RME inventory showing "no device", I'm already getting used to it. It's due to bug CSCef50925 that Cisco is not fixing in RME 3.5.

http://wwwin-metrics.cisco.com/cgi-bin/ddtsdisp.cgi?id=CSCef50925

Displaying Device Lists in RME 3.5 under Configuration Management or Syslog Analysis functions return no records when choosing All Routers or All Devices. Device List is empty. The problem occurs after installing RME IDU 8.0 and is related to database rows getting locked during Inventory Collection.

The only workaround is to stop/restart DIServer and IcServer processes, or stop/restart the CW Daemon Manager, but the problem will resurface again.