I searched this morning for an answer but could not find it. My question is lengthy, due to explanation.
I had a heated conversation the other day with another programmer and a developer who knows almost nothing about DB2, Lf's or almost anything IBM i related. But uses CA PLEX to generate applications and their PF's are on the IBM i.
We are running an IBM i at V7R3, model is a 41a.
The issue was, a rogue LF in my user libe pointing back to a "Production" PF. The LF has not been touched by anything in 7 years, based on the Days used count from the obj desc. Granted, a file in a user libe pointing back to production was an oversight on my part, maybe due to a bad library list, but it was there none the less. The PF is used on a daily bases, but isnt large. About 100 thou records or so.
The argument was, this one lone LF was having a negative impact on something she (the newby) was doing, simply by means of the fact its a logical file and it has to be maintained by the system.
The access path maint parm is default of "*IMMED" and she thought the LF is re-indexed everytime a record in the file is added, changed or deleted.
She asked me to "FIX IT", fix what, the the LF? pointing to the wrong place, delete the LF....WHAT ????
So, when I tried to explain that this LF has no impact to her process or any process currently running, she called out to the other programmer, who said I was wrong and that the LF has to be re-indexed everytime a record is added, changed or deleted. My response was "Not exactly true" due to the access path maint parm of *IMMED and that the system maintains those changes "on the fly" unless the parm is not *immed, like *DLY, *REBLD. And another thing, there are maybe 10-15 LF's over this on PF, due to how CA/PLEX generates them.
Am I absolutely wrong on my understanding of how LF's are maintained and the impact they could have on a system ????
I had a heated conversation the other day with another programmer and a developer who knows almost nothing about DB2, Lf's or almost anything IBM i related. But uses CA PLEX to generate applications and their PF's are on the IBM i.
We are running an IBM i at V7R3, model is a 41a.
The issue was, a rogue LF in my user libe pointing back to a "Production" PF. The LF has not been touched by anything in 7 years, based on the Days used count from the obj desc. Granted, a file in a user libe pointing back to production was an oversight on my part, maybe due to a bad library list, but it was there none the less. The PF is used on a daily bases, but isnt large. About 100 thou records or so.
The argument was, this one lone LF was having a negative impact on something she (the newby) was doing, simply by means of the fact its a logical file and it has to be maintained by the system.
The access path maint parm is default of "*IMMED" and she thought the LF is re-indexed everytime a record in the file is added, changed or deleted.
She asked me to "FIX IT", fix what, the the LF? pointing to the wrong place, delete the LF....WHAT ????
So, when I tried to explain that this LF has no impact to her process or any process currently running, she called out to the other programmer, who said I was wrong and that the LF has to be re-indexed everytime a record is added, changed or deleted. My response was "Not exactly true" due to the access path maint parm of *IMMED and that the system maintains those changes "on the fly" unless the parm is not *immed, like *DLY, *REBLD. And another thing, there are maybe 10-15 LF's over this on PF, due to how CA/PLEX generates them.
Am I absolutely wrong on my understanding of how LF's are maintained and the impact they could have on a system ????
Comment