Translating with DrWatson… this can take a few seconds the first time.
This is a genuine, complex translation. DrWatson protects commands, error codes, and log output while naturally translating the surrounding text. It’s translated once and saved.
John Smith asked whether CLEANERS should equal the number of chunks or disks. Art Kagel said maximum write throughput needs one cleaner per chunk at checkpoint or per LRU queue, ignoring static chunks, and disagreed with IBM manuals. Tom Girsch rarely saw a penalty for too many, but no more than two per spinning disk. Paul Watson usually sets 128 on larger systems.
Auto-generated by Claude from the posts below — may be imperfect; read the full thread.
John Smith — source: IBM Community (ConnectedCommunity.org) Informix forum
Hello
as recommandation, Cleaner parameter should be equal to number of chunks or number of disks ?
Thanks
------------------------------
John Smith
------------------------------
#Informix
↪ replying to John Smith
Art Kagel — source: IBM Community (ConnectedCommunity.org) Informix forum
John:
Whenever buffer flushing is initiated, whether for chunk writes or LRU writes, either a chunk or an LRU queue's dirty buffers are assigned to a cleaner. So, for maximum write throughput, you need one cleaner for each chunk at checkpoint time or one for each LRU queue at LRU flush time.
Now, I note that the maximum number of LRU queues per buffer cache is 512 and you could have multiple caches and that the maximum number of CLEANERS is 128. I would also note that you can discount any chunks that only contain static data when considering the number of chunks.
Obviously I disagree with the recommendations in the IBM manuals, so your decision: one per chunk, one per LRU, one per disk, or one for every 2 disks.
Art
------------------------------
Art S. Kagel, President and Principal Consultant
ASK Database Management Corp.
www.askdbmgt.com
------------------------------
↪ replying to Art Kagel
TOM GIRSCH — source: IBM Community (ConnectedCommunity.org) Informix forum
I've rarely seen a penalty for having too many cleaners configured. The exception would be when you have many chunks on few slower spinning disks (as opposed to SSD), in which case, I wouldn't go more than 2 per disk. Problem is, with modern storage and SANs, you may not actually
know
how many disks there are behind the scenes. In that case, I'd generally follow Art's recommendation of one per frequently-written chunk.
↪ replying to TOM GIRSCH
Paul Watson — source: IBM Community (ConnectedCommunity.org) Informix forum
Like Tom, I have rarely seen any issues with too many CLEANERS. Unless it is a small system I tend to set at 128 and leave it alone
Cheers
Paul
Your privacy choices
We use strictly necessary cookies to make this site work. With your
consent we’d also use optional cookies for analytics and marketing. You can accept all,
reject all, or choose. Read our Cookie Policy.