You are here: Reference Information > Command Line Tools > NuoDB Manager > Remove Database Backoff

Remove Database Backoff

This is a NuoDB Manager command. See NuoDB Manager.

Description

Resets to 0 the number of times that NuoDB tried and failed to restart a process. NuoDB maintains a separate try/fail count for each database process that terminates. This count is referred to as a backoff count because each time NuoDB tries and fails to restart a process then NuoDB backs off (waits) for a given period before trying again. You can reset all backoff counts related to

The backoff.maxRetry database option specifies the maximum number of times to try restarting a database process. If the maximum retry number is reached for any database process then enforcement stops for that database.

To be able to resume enforcement for the database, invoke the remove database backoff command. After that, you can re-enable the enforcer by invoking the NuoDB Manager "Enable Enforcer" on page 1 command or by restarting the database (even if it is already half running) with the NuoDB Manager Start a Database in a Managed Environment command.

Syntax

remove database backoff
    dbname database_name
    [ host hostname ]
    [ group process_group_name ]

Parameters

Interactive Example

The following output shows two processes that are waiting for the backoff delay to elapse so the enforcer can try to restart them. In the UMET report, under Region, Backoff(1) means that the enforcer is waiting to start 1 process. Under Backoff, 1/2 indicates that one of two restart attempts failed and that one more attempt will be made to restart the process before the value set by the backoff.maxRetry database option is reached. Also, you can see which type of processes the enforcer will try to restart on which host and when.

nuodbmgr --broker localhost  --password mypw show database config database foo
NuoDB host version: 3.0: Professional Edition
Database: foo, (unmanaged) ACTIVE, Status=RUNNING, dundant] (v5)
  Variables: {REGION=us-east}
  Options: {backoff.reqMinUptime=60000, verify-unclean-archive=, backoff.stopOnMaxRetry=true, 
      mem=1g, backoff.maxRetry=2, backoff.delay=10000}
  Default Options: { "commit": "${COMMIT:remote:1}","backoff.reqMinUptime":"30000","hostLimit":"${HOST_LIMIT:false}"}
  DB UUID: c44593ce-1c30-2546-869a-affb22b77603
  Process group options:
  Process group tag constraints:
  Archive Locations:
    ip-172-31-34-74/54.88.137.63:48444, requirements: SMs, region: us-east:
      archive: /Users/jdoe/dev/install/nuodb_home2/var/opt/production-archives/foo
      journal-dir: /Users/jdoe/dev/install/nuodb_home2/var/opt/production-archives/foo
    ip-172-31-34-74/54.88.137.63:48004, requirements: SMs, region: us-east:
      archive: /Users/jdoe/dev/install/nuodb_home1/var/opt/production-archives/foo
      journal-dir: /Users/jdoe/dev/install/nuodb_home1/var/opt/production-archives/foo
  
    Region [us-east]:
      process requirement "SMs" (#0): only 1 running SM but 2 are required.  Backoff(1) 
      process requirement "TEs" (#1): only 1 running TE but 2 are required.  Backoff(1) 

    Backoff: 
      [TE] 1/2 ip-172-31-34-74/54.88.137.63:48004  nextRetry: 2015-12-08T16:04:46.427-0500 
      [SM] 1/2 ip-172-31-34-74/54.88.137.63:48004  nextRetry: 2015-12-08T16:04:46.434-0500 

In the following two reports, Backoff [1 SM, 1 TE] indicates the number of processes of each type that are waiting to be restarted.

nuodbmgr --broker localhost --password mypw show domain databases

NuoDB host version: 3.0: Professional Edition
foo [ count = 2 ], (unmanaged), processes [1 TE, 1 SM], ACTIVE, Backoff [1 SM, 1 TE]
bar [ count = 0 ], (unmanaged) STOPPED, processes [0 TE, 0 SM], INACTIVE
nuodbmgr --broker localhost --password mypw show domain summary

NuoDB host version: 3.0: Professional Edition
Hosts:
[broker] * ip-172-31-34-74/54.88.137.63:48004 (us-east) CONNECTED
[broker] ip-172-31-44-72/53.86.14.23:48444 (us-east) CONNECTED

Database: foo, (unmanaged), processes [1 TE, 1 SM], ACTIVE, Backoff [1 SM, 1 TE]

[SM] ip-172-31-34-74/54.88.137.63:51481 (us-east) [ pid = 79834 ] [ nodeId = 2 ] RUNNING
[TE] ip-172-31-44-72/53.86.14.23:51488 (us-east) [ pid = 79836 ] [ nodeId = 4 ] RUNNING

Database: bar, template [Minimally Redundant] UNMET STOPPED

To remove all backoff counts for database foo:

nuodb [domain] > remove database backoff
Database Name: foo
Host or StableId (optional): 
Group (optional): 
removed: backoffStats/foo/uuid:1be18dc6-6d5b-4eb8-9fbc-1ff318c63916/TE/TEs=1
removed: backoffStats/foo/uuid:1be18dc6-6d5b-4eb8-9fbc-1ff318c63916/SM/SMs=1

nuodb [domain] > 

Scripting Example

The same command command can be executed with --command as follows:

To remove all backoff counts for database foo:

--command "remove database backoff dbname foo"

To remove all backoff counts for storage managers that serve database foo:

--command "remove database backoff dbname foo group SM"

To remove all backoff counts for database foo on host localhost:48444 regardless of process type:

--command "remove database backoff dbname foo host  localhost:48444"

To remove all backoff counts for storage managers that serve database foo host localhost:48444:

--command "remove database backoff dbname foo host  localhost:48444 group SM"