Connection between our internal Serial Number with the part's production SN

[xwcwt]

Currently we will record the part in nodelist system and will give a part a Serial No. I think we can call this No as ISN(Internal Serial No.). Also most part has its production Serial No. such as ZAD0KTX of ST6000NM015 disk.

After discussed with Xinjie, in order to make parts recording easilly and save labor time, the part's production SN can be recorded in the thread as the part in such as thread of GM55-1, and we appointed that part with the same Part ID in nodelist sysytem, the smallest ISN maps to the first slot in the node/server. e.g., GM55-1 has 24 ST6000NM0115 disks, the ISN start from SFS20170505 and end with SFS20170538, then the SFS2017050 maps to the first disk slot and this disk's production SN can be gain by showpart.

Once one disk failed and replaced, we update the info in Tab or RAR(thread) in order to make the info consistent , Just imagine the smallest ISN may not point to the first slot if the first disk is replaced with another disk bought later than this batch.

[YU_Xinjie]

Quote:Currently we will record the part in nodelist system and will give a part a Serial No. I think we can call this No as ISN(Internal Serial No.). Also most part has its production Serial No. such as ZAD0KTX of ST6000NM015 disk.

Good to me.

(05-16-2017 05:46 PM)xwcwt Wrote:  After discussed with Xinjie, in order to make parts recording easilly and save labor time, the part's production SN can be recorded in the thread as the part in such as thread of GM55-1, and we appointed that part with the same Part ID in nodelist sysytem, the smallest ISN maps to the first slot in the node/server. e.g., GM55-1 has 24 ST6000NM0115 disks, the ISN start from SFS20170505 and end with SFS20170538, then the SFS2017050 maps to the first disk slot and this disk's production SN can be gain by showpart.

1. "disk slot" is described as "Slot No." by storcli (as well as showpart). Its abbreviation is "Slt".

2. The map of "Slot No." [0, 23] to ISN [SFS20170505, SFS20170538] should be recorded in the thread of node/cluster explicitly.

Quote:Once one disk failed and replaced, we update the info in Tab or RAR(thread) in order to make the info consistent , Just imagine the smallest ISN may not point to the first slot if the first disk is replaced with another disk bought later than this batch.

The idea is good to me. But we need to set up the detailed steps to replace a part, to avoid forgetting update the Slt->ISN info.