The server has really two client ids, as follow:
desenvoet:[kaizen]/tmp> mminfo -av -r clientid,client |grep desenvoet |sort |uniq

0dc16f9a-00000004-45870c7b-45870c7a-00010000-82006464 desenvoet.atacadao.com.br

471d5009-00000004-3ea83471-3ea83470-00010000-82006464 desenvoet

desenvoet:[kaizen]/tmp>

If you access the management console, go to media and click in indexes and visualize the server index, it show the normal index. But when you try a recover from the networker user and choose a data, for example, 12/6/2007, or the SSID, it says that it doesn¿t have backup, showing the following error:

No entries backed up to server desenvoet.xxxxx.com.br at time 6-12-07 ( recover via Netwoker user GUI )

recover> changetime 06/11/07
cannot browse /tmp/ as of ` 6/11/07 23:59:59' ( recover via console CLI )
recover>

Every time that a new backup is made, the index created by the previous backup is deleted, but as I said before, I can visualize the index in the management console. Apparentely, it is reseting always from the last index, after the last backup of the day is made. In the NW index checking it seems that it is consulting a client Desenvoet before it consults the desenvoet.atacadao.com.br ( the correct name). I don¿t know wheter it influences or not in the problem that is occurring.