Raj.la,

Can you give "isi config >>>status advanced "output?

Example:

s200-2# isi config

...

s200 >>> status advanced

Configuration for 's200'

...

Device health:

LNN  | ID  | Cluster IP      | GUID                                | Type        | Status |

-----+-----+-----------------+--------------------------------------+-------------+--------+

1    | 1* | 128.222.251.1   | 000743092dc012c5c5573f194a002f8c5cxx | Storage    | Okay  |

2    | 2   | 128.222.251.2  | 000743093c1074c8c5571111040dba6497xx | Storage    | Okay  |

3    | 3   | 128.222.251.3  | 000743092ccc59ccc557340822d7386036xx | Storage    | Okay  |

-----+-----+-----------------+--------------------------------------+-------------+--------+

s200 >>>

...

That's the node ID (not the LNN, which is different) being appended to the hostname.  The node ID, unlike the LNN, is the never-recycled identifier that uniquely represents a node.  The syslogd source was modified I think back in the OneFS 7.2.1 days to include this.

I'm sure this was done to enable more accurate parsing of logs, since hostnames and LNNs can change.  That makes sifting through logs tricky unless there's some type of uniqueness to the string.

Hi Raj.la,

To expand one what has already been said, each node has a logical node number (LNN) and a node ID. Like kipcranford said, the node ID does not get recycled, but the LNN does.

Here is an example of how LNN prod-isilon-6 could become id9

Initial Cluster Config

Node name       LNN     ID
prod-isilon-1     1          1

prod-isilon-2     2          2

prod-isilon-3     3          3

prod-isilon-4     4          4

prod-isilon-5     5          5

prod-isilon-6     6          6

prod-isilon-7     7          7

prod-isilon-8     8          8

Node 6 has a problem with its motherboard and a chassis swap is completed. Essentially this means the node has been replaced.

Config after chassis swap

Node name       LNN     ID
prod-isilon-1     1          1

prod-isilon-2     2          2

prod-isilon-3     3          3

prod-isilon-4     4          4

prod-isilon-5     5          5

prod-isilon-6     6          9

prod-isilon-7     7          7

prod-isilon-8     8          8

As you can see, the LNN stayed the same. It is still the 6th node in the cluster, but the ID number changed because this node has a new serial number, motherboard, etc. There are other things that can cause the ID number to be different from the LNN (renaming LNN's or splitting up a cluster for example), but most common is a node replacement.

A quick note on this example: I appears that node 6 has been smartfailed or stopfailed before the replacement, and then the replacement node had to be completely repopulated with data (by MultiScan), therefore a new node ID was necessary.

It is also possible to swap chassis and motherboard while keeping all disk content intact -- which doesn't not require a  new node ID, and MultiScan finishes quickly like after a plain reboot. This is usually the more desirable kind of hw replacement and it should not be confused with the first case (which of cause was chosen here to demonstrate an actual change of the node ID).

fwiw

-- Peter