HiPath 4000 Backup Error / Database inconsistency 4

[lhiraman]

I just noticed that we are unable to get a backup of our Hi-Path 4000 V5 backup for a week now. There is an errors:
Error in pre-/post-backup/restore script.[39]
Can't make backup set.[14]

I can't seem to figure it out. I have a ticket open with Siemens for about 3 days now.

I was changing a name on a phone today, and when I wanted to save via assisant, I get a database consisentency error (try a upload).

I HAVE NO CLUE! Does anyone know how to fix this?

 

[Iamnothere]

Did you try to perform an Upload? Before I explain, you should NOT perform the following tasks during heavy traffic on the 4K; but rather wait until after hours when there is little to no traffic!!
When you perform Moves, Adds, Changes via AMO on a HiPath 4000, you are writing directly to the active Switching Unit memory. The contents of this memory should be frequently copied to the Hard Drive using AMO UPDAT, usually performed on a schedule (see DIS-UPDAT). Assistant also needs a copy of this database, so it uses AMO UPLO2 to fetch a copy each evening. Occasionally the three copies of the database (SWU, Hard Drive partition E, Assistant/Informix) can get out of sync, meaning that the three copies are not "equal". You must assume that the database within the active Switching Unit is always accurate; therefore perform a EX-UPDAT:BP,ALL; twice, and then a EX-UPDAT:A1,ALL; twice to update the hard drive partition E. Next, manually trigger an "upload" from Assistant by clicking Configuration Management > Network > System, and perform a Search. There you should see the status of the Informix database within Assistant. If you see "Synchronized" in each and every component of the database ("Stations" for example), then typically no upload is necessary. (In your situation, regardless of the statuses, you should start an Upload All.) For any other status, start an Upload All by clicking "Action" on the upper action bar, then click "Upload". If prompted whether to run an Upload All or Upload Delta, you should select Upload All in your situation. When done, you should see "Synchronized" in all of the database component fields. Then you might see the overall status of "Backup Required", which means the database needs to be saved to the hard drive (AMO UPDAT), even though you may have performed this task manually already. To satisfy Assistant, you should perform a "SAVE" from this same window, again by clicking "Action" on the upper action bar, then click "SAVE". The status should update to "Synchronized" when done.

This may not resolve the problem - you could have a more serious database problem. SEN may have to perform a GENDB, which means that your database is re-generated. Usually this action resolves most database problems. Must be done during OFF-HOURS. System will be down during this process.

Please re-post if/when problem is resolved, OR if I have mis-judged the problem. Good luck.

 

[lhiraman]

Ok the "Configuration Management > Network > System, and perform a Search" seem to have taken care of the consistency database error. I will proform the other AMOS task after hours. A siemens tech is supposed to be here today.

 

[lhiraman]

I just spoke with the Siemens tech, he said that he needs to reboot/ restart the A1 card. As a first step to troubleshooting the backup error.

 

[lhiraman]

Ok The Siemens tech is telling me that they have to conduct a UNIX recovery to get the backups to run again.

I am concern about losing any data, he said that everything will be intacted. WHAT should I be asking and how safe is this?

Thank in advance!

 

[Iamnothere]

Unix recovery is not a big deal. Unix is merely the operating system for the administrative side of the HiPath 4000. If it has become corrupted, then it cannot adequately process its copy of the system's database. The tech will simply re-install Unix from the RMX partition "I", add all the necessary hotfixes, and your Assistant will be operational again within a few hours. The system will not have to be down to perform the Unix recovery. Is your tech a Siemens employee, or a partner/distributor?

 

[lhiraman]

Thanks Iamnotthere!

The tech is a Siemens employee, when I spoke with him, he said that our Procenter will be down and it will take him about 3 hours to do. He was waiting on documentation from BLS or BSL something like that. My manager said that this work should be done on a weekend, but the Siemens tech wants to do it on May 9.

Also, We conducted a system back of RMX and the UNIX portion to a FTP SERVER; can i restore only the unix portion or just relax and let the tech so the recovery?

 

[Iamnothere]

The tech performed the backups now so that after he re-installs Unix, he can re-configure FTP access to your external backup server, and "restore" Unix from today's backup. The restore will re-install your custom-created userIDs for Assistant, etc - all things specific to system management. The backup also contains a copy of today's database, in the unlikely event that your system crashes during the Unix recovery - on May 9 he can restore the RMX database to the state that it is in today. The HiPath 4000 does not need Unix to provide dial tone, as it is an administrative/management operating system only. Therefore the likelihood of your system "crashing" during the recovery is extremely slim. He's contactling SEN Back Level Support for specific instructions on how to recover Unix on your system.

Sounds like you are in good hands. Whoever the tech is, there is a 99% chance he knows me, regardless of his location. But unfortunately I cannot identify myself here. I cannot even mention why. That's why my screen name is 'I am not here' !!!!!

 

[lhiraman]

Iamnothere, thank you! you are making me feel more and more comfortable! Don't know if I can post the tech that is responsible for our site.

THANK YOU!!! You have been a great help, since I am not a Siemen tech just a regular IT person who this role was just giving to.

 

[donb01]

Sounds like everything is under control, but one side note about your ProCenter. Procenter is basically responsible for all of the reporting and statistical functions for your call center, and when it is running it handles all of the call routing, etc.

Normally your system will also be configured so that if ProCenter fails is has backup routing in place so that all your agents can still login and will get their calls; you just won't get any reports or statistics for the time ProCenter is down, and you would not get any messages that are driven by ProCenter such as "Your average hold time is XX minutes" and things along those lines. Customer calls should still get queued, they should still get hold music if you provide it, and they should get some basic announcements if your system is configured with backup announcement devices.

The switch should be able to detect that ProCenter is down and it will normally switch to the backup plan automatically, so your call flow shouldn't be interrupted there either.