Nonupdate error with s flag on while querying paritioned hdbs

Community Support
1

any example to get the below nonupdate error?

To avoid noupdate errors on queries on partitioned tables, put count table in your startup script.

2

The first time a query runs a select count i style query a cache is populated of rows per date. If this is tried inside a thread using peach and the process has treads avilable (.e.g -s 2) it will error with noupdate as globals cannot be updated from within threads.

$ q -s 2

q)\l HDB
q)tables[]
`s#`test`test2
q){select count i by date from x} peach tables[]
'noupdate: `.Q `PN
  [6]  (.Q.cN)

  [5]  (.Q.cn)

  [4]  (.Q.dt)

  [3]  (.Q.ps)

  [2]  {select count i by date from x}
        ^
  [0]  {select count i by date from x} peach tables[]
                                       ^
q))\

Without threads peach acts the same as each so the .Q.pn cache is populated successfully.

$ q

q)\l HDB
q)tables[]
`s#`test`test2
q){select count i by date from x} peach tables[]
(`s#+(,`date)!,`s#2026.02.10 2026.02.11)!+(,`x)!,3 3
(`s#+(,`date)!,`s#2026.02.10 2026.02.11)!+(,`x)!,3 3
q).Q.pn
test | 3 3
test2| 3 3