Hello, Just like every connection, may be you can find such message in dmesg/logs when IPVS starts: IP_VS_DBG(0, "Each connection entry needs %Zd bytes at least\n", sizeof(struct ip_vs_conn)); You ca
Hi, I'm trying to calculate how much memory will be needed to synchronize the persistent templates across the entire cluster. How much memory consumes one persistent template? Aleksey -- To unsubscri
Hello, On 19.06.2013 12:03, Julian Anastasov wrote: Here is something that is compile-tested. You will need the "ipvs: sloppy TCP and SCTP" patch by Alexander Frolkin posted on 13 Jun. Let me know if
Hello, Ops :) s/in_peristence/in_persistence/ Regards -- Julian Anastasov <ja@xxxxxx> -- To unsubscribe from this list: send the line "unsubscribe lvs-devel" in the body of a message to majordomo@xxx
Hello, Here is something that is compile-tested. You will need the "ipvs: sloppy TCP and SCTP" patch by Alexander Frolkin posted on 13 Jun. Let me know if you need more help in applying and testing s
Hello, You can also fix this comment. After fixing above comment and if tests looks ok you can post an official patch. Regards -- Julian Anastasov <ja@xxxxxx> -- To unsubscribe from this list: send t
Hello, I don't see problems with this version. If you post an official patch I'll ack it. Others still have time for comments. Regards -- Julian Anastasov <ja@xxxxxx> -- To unsubscribe from this list
Hi, Latest version of SH patch: diff --git a/include/uapi/linux/ip_vs.h b/include/uapi/linux/ip_vs.h index a245377..2945822 100644 -- a/include/uapi/linux/ip_vs.h +++ b/include/uapi/linux/ip_vs.h @@
Hello, If not a multi-line, the comment can be: /* Helper function to determine if server is unavailable */ But some function names are too obvious :) No need for extra 'else' May be 'iph', not '&iph
Hello, Such strncpy calls are not very safe, they should be strncpy(dest, src, space - 1); dest[space - 1] = 0; Better to use snprintf(sched_flags_arg, sizeof(sched_flags_arg), "%s", optarg); We can
Hi, I've moved things around a bit: diff --git a/include/uapi/linux/ip_vs.h b/include/uapi/linux/ip_vs.h index a245377..2945822 100644 -- a/include/uapi/linux/ip_vs.h +++ b/include/uapi/linux/ip_vs.h
Hi, I've moved the code to parse the scheduler flags to the end of the function, when the scheduler is known: diff --git a/ipvsadm.8 b/ipvsadm.8 index 001ae74..9a9e9b3 100644 -- a/ipvsadm.8 +++ b/ipv
Hello, Wait, it seems is_unavailable() can succeed for other reasons, so the 'if (!dest) ... break;' check should be first, may be out of is_unavailable(). Regards -- Julian Anastasov <ja@xxxxxx> --
Hello, Yes, you will find 256 NULLs there. Regards -- Julian Anastasov <ja@xxxxxx> -- To unsubscribe from this list: send the line "unsubscribe lvs-devel" in the body of a message to majordomo@xxxxxx
Hi, So if dest is NULL, there is no point trying to choose a different server? Alex -- To unsubscribe from this list: send the line "unsubscribe lvs-devel" in the body of a message to majordomo@xxxxx
Hello, OK, if others do not come up with another idea. sched-flags here? TAG_SCHED_FLAGS is not needed with 'b' Moving the following parsing in separate function parse_sched_flags(sched, optarg) will
Hello, There is already a requirement the multiline comments in net/ to be in such format: /* First line * ... * last line */ Do it for all comments that you add in this patch. 'bool' should work eve
Hi, Updated ipvsadm patch below. I've kept -b for now: there are no long options in ipvsadm -S output, and I think it would be nice to keep it that way. diff --git a/ipvsadm.8 b/ipvsadm.8 index 001ae
Hi, Well, okay. I just have a personal preference for short options. :-) It makes ipvsadm -Sn output look nicer, too. Okay, that looks good. I've done some refactoring to simplify ip_vs_sh_schedule.