Slow performance after shmem dynamically allocated

_______________________________________________________________ This message and any attachments are intended for the individual or entity named above. If you are not the intended recipient, please do not read, copy, use or disclose this communication to others; also please notify the sender by replying to this message, and then delete it from your system. Thank you. _______________________________________________________________ I've heard that there was a bug in previous versions of IDS that performance suffered after a new Virtual Memory segment was dynamically allocated. We're using 7.31.UC6 on HP 10.20. Does anyone know of any problems with that in this version? TIA. BTW, please respond directly to this address, thanks. Colleen

Just a minor correction is in order here. The problem that you are refering to is not a bug in IDS. It is a problem that any software will encounter if running on HP and is allocating shared memory segments. The problem is actually caused by the way that HP manages shared memory segments and utilizes the space registers. Basically when more than four memory segments are constantly being referenced, then there will be performance degredation. This is because HP actually only bases four segments at any given time. If there are actually five, then when an access to a memory address is made to a segment that is not currently based, then a check is made to see if it *could* be resolved by one of the un-based segments. If it could, then that segment's basing information is loaded into one of the space registers and the access is retried. Needless to say, this could cause a bit of overhead. OK - lets look at where the space registers are used. One is always basing the text portion of the program. A Second is basing the shared object library. That leaves two for shared memory. But there are three main shared memory portions that are going to be constanly referenced - 1) the buffer pool, 2) the dictionary, 3) the user stack (i.e. multi threaded information). We've tried to minimize the impact of this HP characteristic by combining the buffer pool and the first virtural memory segment into a single physical segment. That tends to minimize the impact quite a bit. However it is still advisable to make the fixed virtural memory segment significantly larger on HP systems so that the possibability of thrashing between the memory segments is diminished. "Colleen M. Morrow" wrote: > _______________________________________________________________ > > This message and any attachments are intended for the individual or entity > named above. If you are not the intended recipient, please do not read, > copy, use or disclose this communication to others; also please notify the > sender by replying to this message, and then delete it from your system. > Thank you. > _______________________________________________________________ > > I've heard that there was a bug in previous versions of IDS that > performance suffered after a new Virtual Memory segment was dynamically > allocated. We're using 7.31.UC6 on HP 10.20. Does anyone know of any > problems with that in this version? TIA. > > BTW, please respond directly to this address, thanks. > > Colleen

It's actually an HP-UX 10.20 issue. Try to keep single sgared memory extent if you possibly can (even at the cost of wasting memory in the initial allocation), or die slow performance death! Colleen M. Morrow <cmmorrow@JonesDay.com> wrote in message news:8rcrdi$de7$1@news.xmission.com... > > > _______________________________________________________________ > > This message and any attachments are intended for the individual or entity > named above. If you are not the intended recipient, please do not read, > copy, use or disclose this communication to others; also please notify the > sender by replying to this message, and then delete it from your system. > Thank you. > _______________________________________________________________ > > I've heard that there was a bug in previous versions of IDS that > performance suffered after a new Virtual Memory segment was dynamically > allocated. We're using 7.31.UC6 on HP 10.20. Does anyone know of any > problems with that in this version? TIA. > > BTW, please respond directly to this address, thanks. > > Colleen >