Avamar GEN4 Visio stencil

Looking for Gen4 hardware stencils ?

Just use DELL's stencils.
The utility node is a Dell R710 server , and the storage nodes are R510 servers.

Symmetrix V-Max head / track / cylinder conversion

Symmetrix V-Max
Enginuity Code: 5874
Includes EMC Symmetrix V-Max and various different configurations within this model.
GB = Cylinders * 15 * 128 * 512 / 1024 / 1024 / 1024
Eg: 262668 Cylinder device equates to 240.47 GB of usable data
262668 * 15 * 128 * 512 / 1024 / 1024 / 1024 = 240.47 GB

Cylinders = GB / 15 / 128 / 512 * 1024 * 1024 * 1024
15 = tracks per cylinder
128 = blocks per track
512 = bytes per block
8 bytes = 520-512 used for T10-DIF
1024 = conversions of bytes to kb to mb to gb
Drive format on a V-Max is 520 bytes, out of which 8 bytes are used for T10-DIF ( A post on DMX-4 and V-Max differences).


(Thank you storagenerve.com)

Avamar 7.0.1 - Disabled groups are being reported as failed

Since the last version of Avamar, any clients located within a disabled backup group are now
being reported as "Failed" in the daily  summary. This is good if for some reason you forgot to re-enable a group after a temporary exclusion, but very annoying if you have groups disabled for some good reasons.

This behavior can be change in the mcserver.xml file ( Please be careful ):

1. Make a backup of mcserver.xml just in case.

cp -p /usr/local/avamar/var/mc/server_data/prefs/mcserver.xml /usr/local/avamar/var/mc/server_data/prefs/x-mcserver.bak20140923

2. Edit the mcserver.xml file

vi /usr/local/avamar/var/mc/server_data/prefs/mcserver.xml

3. Change this key from true to false

<entry key="report_disabled_client_job_as_fail" value="false" />

4. Restart MCS and the backup scheduler

dpnctl stop mcs; dpnctl start mcs; dpnctl start sched

Avamar error - Error writing 32-byte header to cache file

You most likely have "avtar.exe" process stucks. This seem to be happenning the most often on MSSQL clients.

Kill them, and restarts Avamar services.

Avamar - MCCLI - 1,22801,User login failure.

After running /usr/local/avamar/7.0.1-61/bin/avsetup_mccli for the first time, if you make an error in your setup, it will never go back and update files in your home ~/.avamardata/* correctly.


Fix:

1. delete ~/.avamardata/* OR:
2. manually update ~/.avamardata/7.0.1-61/var/mc/cli_data/prefs/mcclimcs.xml with the correct information.

How to suppress errors and backup exceptions in Avamar Activity and enterprise manager

So you don't want to know about a certain error, or the administrator of that system is not helping you solve this backup problem ?

No problem, just do the following so it will not appear in your "failed" list:

Go to the AVAMAR_INSTALL_PATH\avs\var directory 
Create a file named avtar.cmd

Append the following in it (replace 30901 with the error code you want to suppress):

--downgrade-error-codes="30901"

Avamar Grabage collection troubleshooting

If your garbage collection stop working, you know you are in trouble. Your grid will eventually fill up, and die a slow and painfull death.
Use the following commands from the utility node to try and diagnose the problem:

dumpmaintlogs --types=gc --days=7
In case you see along the lines of "<4203> garbage collection skipped because backups in progres", please verify which nodes are allowed overtime, they are certainly creating problems.

Nodes that are allowed overtime can be seen with:
mccli client show --recursive=true --verbose=true | grep -v "No overtime allowed"


If you'd rather just have garbage collection kill the sessions preventing it from running, just type:
avmaint sched gc --ava --kill=300 --permanent

gc_report
less /usr/local/avamar/var/cron/gc.log

Avamar - How to retrieve all of your grid's serial numbers

Sometimes, EMC support will contact you because your equipment dialed in, and need a part replaced. They will mention a node serial number, so if you have multiples sites like me, here is a useful command to retrieve your grid's serials:

Log in to the utility node:

#> su - admin
#> ssh-agent bash
#> ssh-add ~/.ssh/dpnid
#> mapall --noerror --user=root --nodes=all+ --quiet 'echo `hostname -i`": "`/usr/bin/ipmitool fru print 0 | grep "Product Asset Tag" | sed "s/^.*:\s\(.*$\)/\1/"`'

This will get you the node serial number only.
To get more extensive information, you can try this as well:
#> mapall --noerror --user=root --nodes=all+ --quiet 'echo `hostname -i`": "`/usr/bin/ipmitool fru print 0 " grep "Product Asset Tag" " sed "s/^.*:\s\(.*$\)/\1/"`'

(see solution esg111183 on emc support website)