DRS back up fails

[jidesh chandran]

while doing the Disaster Recovery System back up getting error

ERROR: SFTP transfer failed as backup size was not increasing for the past 15 minutes. Either there is not enough disk space or network transfer rate is too slow with the configured SFTP Server. Please either free some space on SFTP device or check network connectivity and then run a fresh backup, Backup Completed...

in this cucm cluster 5 servers are available ,all the first three where taken back up 100% once start the back up on the 3 rd one  MOH,ANN,CEF,TCT,CDPAGT,REPORTER taken 100% but when it start the TFTP it is not showing the percentage but able to see that increasing the file size on the sftp server but not showing on web interface of DRS and after 15 min the back up canceled automatically . as we checked the there is no lan restriction between the sftp server and CUCM .

current version of the CUCM is  8.6.2.22900-9

[Aman Soi]

Hi Jidesh,

Can u share the snapshot along with logs where it is getting stuck on TFTP for 3rd server?

regds,

aman

[Jagpreet Singh]

Hi Jidesh,

     Please restart the DRF Master on the Publisher and the DRF Local on the SUB having the issue. Please also restart the TFTP service on the SUB having the problem.

Also, what is the SFTP client which we are using to take the backup?

Regards,

Jagpreet

[Pratyushm]

Hi All,

I have the same problem with CUCM 8.6.2.20000. The BAT componenent completes successfully but the when it comes to the TFTP component the backup fails with:

ERROR: SFTP transfer failed as backup size was not increasing for the past 15 minutes. Either there is not enough disk space or network transfer rate is too slow with the configured SFTP Server.

i have used freeftpd and coreFTP but both giveme the same error.

thanks in advance.

-

Cheers,

Pratyush

[Aman Soi]

Hi Manjul,

How many servers do you have in a cluster?

On how many servers TFTP is running  and tfp fails on which server while taking backup.

Can u share the snapshot from DRS which shows error?

regds,

aman

[Pratyushm]

There are only 2 servers in the cluster and tftp is running on both. The problem happens when backing up the publisher.

Thanks.

-

Pratyush

[Aman Soi]

Hi,

This means the backup stops at TFTP Publisher.

There is a possibility that no. of files are more in  tftp in Publisher than Subscriber.Can u share the logs by clicking on the TFTP in DRS while it stops.

Can u try stopping the TFTP service in Publisher and then, take backup?

regds,

aman

[Jean-Christophe Majchrzak]

Hi Jidesh,

I'm facing the same problem.

Running FreeFTPd soft v.1.0.11 on a Win2008 server R2 Standard same issue as yours.

Running the same soft, same version on old Win2003 server SP2, it works perfectly.

JC,

[antonypullella]

Hi

I am having the same issue. It is as if one of the servers thinks it is idle for 15mins and then stops the transfer.

I only have this problems since moving from Server 2003 running freeFTPd 1.0.11 to Server 2012 running freeFTPd 1.0.12.

The TFTP file was able to hit 700-900mb during the day before failing and over night it hit just over 1GB.

I need to backup 1.5gb based on my old backups. Maybe some issue between running newer version of Server, I have the firewall off.

Any help would be great.

Thanks

Antony

[Jagpreet Singh]

HI Jidesh,

     Any update on this post?

Hi Antony/Jean,

     Freeftpd is not really Cisco recommended for backups. We recommend :

open SSH

Core FTP

Cygwin

Titan

Hi Pratyush,

As Aman requested, could you please share the DRF Master and Local logs for a failed backup from the servers. Also have you tried to restart the DRF services?

Regards,

Jagpreet

[Pratyushm]

Hello Everyone,

My apologies for replying late.

For me the problem was on the LAN. One of the physical LAN connections was working only at 10M half duplex despite setting it up for AutoNeg on both sides and that link incidently was acting as the primary in the bundle. Once we removed it from the bundle the backup worked like a charm.

Thanks everyone for the time and effort.

Much appreciated.

-

Cheers,

Pratyush

P.S. - Do check you LAN connections, a Half duplex connection may not cause protocols like Http to degrade but it wil more often than not cause the FTP connection to become extremely slow and eventually drop.

[Aman Soi]

Thanks for update.

regds,

aman

[Alexandra Costea]

Hi!

I am facing the same issue. CUCM version:

System version: 8.6.2.22900-9

Error appears on both PUB and SUB. We are using Titan FTP.

Error message:

ERROR: SFTP transfer failed as backup size was not increasing for the past 15 minutes. Either there is not enough disk space or network transfer rate is too slow with the configured SFTP Server. Please either free some space on SFTP device or check network connectivity and then run a fresh backup, Backup Completed...

./pkg1.ciusSP.9-2-3-143.tgz
tar: -: Cannot write: Broken pipe
tar: Error is not recoverable: exiting now
Backup failed

This is only happening for the US and Brazil clusters, which are both running the same system version. And for the CUCM, Unity backs up with no errors.

Is anyone aware of a possible resolution? It doesn't seem to be network-related, but we having exactly ignored that possibility. We believe that it might not be network because the Unity servers at those location are not facing the same issue.

Any tip will be very much appreciated. Thank you very much!

[Aman Soi]

Hi Alexandra,

Can u please look into this bug

CUCM DRS backup fails with windows 7 and 2008

CSCul38509

Description

Symptom:
CUCM 8.6+ DRS backup may fail due to interoperability issues with Windows-based SFTP servers.

Conditions:
This has been observed under Windows 7/2008.

To check if you are affected please conduct the following steps during file upload to your Windows-based SFTP server:

a) open CMD and change your current folder to SFTP root directory,

b) issue dir command to check what is the size of the file being uploaded,

c) you are affected if the file size is 0, it is not increasing, and it gets updated as soon as you issue dir /R command.

Workaround:
Use linux-based SFTP server.

Further Problem Description:
https://supportforums.cisco.com/thread/2218075 . This arises due to fix implemented with CSCtc49288

regds,

aman

[Alexandra Costea]

Hi Aman,

You are right, this is exactly what we are experiencing.

Thank you very much for your help!

KR,

Alex