>> If you already are, are you really doing enough writes to >> swamp a NFS cache server? It probably requires hundreds of >> compiling clients; since we have over a hundred here and >> don't see a bottleneck - with a single well performing NFS >> server. > >In our case we're talking about 25 shared builds coming from like 15 >machines. >Can you disclose what the specs are on that machine?
We have about 100 machines, with about 400 cores - a mix of older 2 socket up to modern 12 core systems. Compiles are ~~10% of the work load. The NFS server with CCACHE_DIR has 6 RAID-5 disks; it does other things too, not just ccache. We decided to bond 4 ethernet ports into the NFS server as we saw occasional bottlenecks exceeding 1GbE out of that server. I should have also mentioned that there's also a ccache patch set I've inherited that allows two level caching (local disk, then to the remote disk), that would further improve performance, but we're not currently using it. >> Memcached would provide a nice benefit of providing >> tolerance for machines going down, and somewhat better >> latency, but perhaps the above ideas with the existing >> version can deliver enough performance for you. >> > >I think the idea I like most about this is the simplicity and less >overhead. We're talking about plain tcp sockets without configuration >needed. Managing CIFS/NFS isn't a big deal mostly because most IT departments already know how to manage CIFS/NFS. Also, I'd be a little surprised if your site doesn't have shared CIFS drives already. >I see that memcached is limited to 1 mb data per key. Naturally this >causes some troubles as many files would either not be cached or you'd >need to split it up to more keys. Good point, forgot about that. -Wilson _______________________________________________ ccache mailing list email@example.com https://lists.samba.org/mailman/listinfo/ccache