Re: [harbour] Digest for harbour-devel@googlegroups.com - 5 updates in 2 topics

[matt johnson]

Maybe DNS ( Domain Name Server ) error, it is waiting for timeout, then going to secondary DNS server? 

On Monday, January 22, 2024 at 06:48:07 PM CST, harbou...@googlegroups.com <harbou...@googlegroups.com> wrote:

harbou...@googlegroups.com

Google Groups

Topic digest
View all topics

• NETIO_MTSERVER on new hardware is much slower than expected - 4 Updates
• 2023-12-08 23:17 UTC+0100 Phil Krylov (phil a t krylov.eu) - 1 Update

NETIO_MTSERVER on new hardware is much slower than expected

Lorenzo Fiorini <lorenzo...@gmail.com>: Jan 22 07:12PM +0100 Hi all,   I'm trying to upgrade an installation that is using a Linux server with Ubuntu 18.04.5 LTS 32bit since years as a NETIO_MTSERVER with a newer hardware with a Debian 12 64bit but the new setup is much slower than the old one (5-6 times slower).   It's not a hardware or a network problem since I get the same bad result in my office even with a much more powerful hardware.   I've tried to change several sysctl parameters and I've also tried to use the very same NETIO_MTSERVER 32bit code that I use in old Ubuntu without success. As a comparison I've tried to install Samba and set a simple share and it seems quite fast, surely faster than the old setup and much much faster than the new setup.   Any suggestions?   Regards, Lorenzo

Gerald Drouillard <ger...@drouillard.ca>: Jan 22 01:18PM -0500 Tell us more about what the netio server is slow at. example: dbf access?   On Mon, Jan 22, 2024 at 1:12 PM Lorenzo Fiorini <lorenzo...@gmail.com> wrote:

Lorenzo Fiorini <lorenzo...@gmail.com>: Jan 22 09:10PM +0100 On Mon, Jan 22, 2024 at 7:18 PM Gerald Drouillard <ger...@drouillard.ca> wrote:   > Tell us more about what the netio server is slow at. example: dbf access?   It is a very simple Harbour app, that I have used almost untouched for years, that "serve" a folder of dbf/cdx files via a TCP port using pretty standard NETIO_MTSERVER() code. Windows clients connect to it using NETIO_CONNECT() (see contrib/hbnetio). This client/server setup has been working flawlessly for years. What I wanted was simply faster data access.

Gerald Drouillard <ger...@drouillard.ca>: Jan 22 03:57PM -0500 Did the file system change? Disk r/w rates? Test network I/O against the new server? Memory, cpu usage? caching policy Look in the kernel, system logs? What is slow? locking? opening? seeks? Does the size of the file matter? Are the dbf's also available via samba? Maybe you have a windows machine virus scanning the file(s) on access?     On Mon, Jan 22, 2024 at 3:10 PM Lorenzo Fiorini <lorenzo...@gmail.com> wrote:

Back to top

2023-12-08 23:17 UTC+0100 Phil Krylov (phil a t krylov.eu)

Phil Krylov <phil....@gmail.com>: Jan 22 02:49PM +0100 Hi Luigi,   > ../../../hbpp.c: In function 'hb_pp_generateVerInfo': > ../../../hbpp.c:353:22: warning: format '%u' expects argument of type 'unsigned int', but argument 3 has type 'HB_MAXINT' {aka 'long long int'} [-Wformat=] > 353 | fprintf( fout, "#define HB_VER_REVID %" PFHL "ulu\n", nRevID );   Thanks! To be able to reproduce the issue, one would need to know if you're building from the latest Harbour source (assumed yes) and what version of MinGW-W64 runtime (not just the gcc compiler but the headers) is used. Also, the actual failing gcc command llne would be helpful. On the GitHub CI, with rolling-release MSYS2, the warning does not reproduce.   This is a change probably related to your issue: https://github.com/harbour/core/pull/333/files   -- Ph.

Back to top

You received this digest because you're subscribed to updates for this group. You can change your settings on the group membership page. To unsubscribe from this group and stop receiving emails from it send an email to harbour-deve...@googlegroups.com.