当前位置:   article > 正文

iptables使用limit控制新建连接速率_iptables limit

iptables limit

如下limit匹配的帮助信息。

# iptables -m limit -h

limit match options:
--limit avg                     max average match rate: default 3/hour
                                [Packets per second unless followed by 
                                /sec /minute /hour /day postfixes]
--limit-burst number            number to match in a burst, default 5
  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7

首先新建一个链,将连接跟踪新建状态的报文,导到RATE-LIMIT链中。

# iptables --new-chain RATE-LIMIT
# iptables --append INPUT --match conntrack --ctstate NEW --jump RATE-LIMIT
  • 1
  • 2

以下在RATE-LIMIT链配置规则,使用limit将报文速率控制在每秒50个以下,由于此链中的报文都是连接新建报文,即将连接新建速率限制在每秒50个。超过此速率的报文,执行丢弃DROP。

# iptables --append RATE-LIMIT --match limit --limit 50/sec --limit-burst 20 --jump ACCEPT
# iptables --append RATE-LIMIT --jump DROP
  • 1
  • 2

limit匹配

由函数xt_register_matches注册匹配结构limit_mt_reg。

static struct xt_match limit_mt_reg __read_mostly = {
    .name             = "limit",
    .revision         = 0,
    .family           = NFPROTO_UNSPEC,
    .match            = limit_mt,
    .checkentry       = limit_mt_check,
    .destroy          = limit_mt_destroy,
    .matchsize        = sizeof(struct xt_rateinfo),
#ifdef CONFIG_COMPAT
    .compatsize       = sizeof(struct compat_xt_rateinfo),
    .compat_from_user = limit_mt_compat_from_user,
    .compat_to_user   = limit_mt_compat_to_user,
#endif
    .usersize         = offsetof(struct xt_rateinfo, prev),
    .me               = THIS_MODULE,
};
static int __init limit_mt_init(void)
{
    return xt_register_match(&limit_mt_reg);
  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19

配置检测函数如下,突发burst的值不能为零,平均值avg和突发burst的乘积不能小于平均值avg,即burst大于等于1。

static int limit_mt_check(const struct xt_mtchk_param *par)
{   
    struct xt_rateinfo *r = par->matchinfo;
    struct xt_limit_priv *priv;
    
    /* Check for overflow. */
    if (r->burst == 0
        || user2credits(r->avg * r->burst) < user2credits(r->avg)) {
        pr_info_ratelimited("Overflow, try lower: %u/%u\n", r->avg, r->burst);
        return -ERANGE;
    }
  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11

规则配置的avg与burst的乘积转换为最大的可用信用值。如果cost为空,将credit_cap初始化为最大信用值,将cost初始化为平均信用值。

    priv = kmalloc(sizeof(*priv), GFP_KERNEL);
    if (priv == NULL)
        return -ENOMEM;
    
    /* For SMP, we only want to use one set of state. */
    r->master = priv;
    /* User avg in seconds * XT_LIMIT_SCALE: convert to jiffies * 128. */ 
    priv->prev = jiffies;
    priv->credit = user2credits(r->avg * r->burst); /* Credits full. */
    if (r->cost == 0) { 
        r->credit_cap = priv->credit; /* Credits full. */
        r->cost = user2credits(r->avg);
    }
  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13

函数user2credits将用户配置值转换为内核使用的信用值,应用层下发的均值avg,单位使用的是:(seconds * XT_LIMIT_SCALE),换算为对应秒值需要执行除法操作:avg/XT_LIMIT_SCALE。所以将用户配置的均值avg转换为信用值credits,需要再次乘以每秒对应的credits数量,即HZ*CREDITS_PER_JIFFY。

在转换之前,检查溢出情况,如果三者(user * HZ * CREDITS_PER_JIFFY)乘积大于32bit的最大值,先除后乘;否则,先乘后除。

/* Precision saver. */
static u32 user2credits(u32 user)
{
    /* If multiplying would overflow... */
    if (user > 0xFFFFFFFF / (HZ*CREDITS_PER_JIFFY))
        /* Divide first. */
        return (user / XT_LIMIT_SCALE) * HZ * CREDITS_PER_JIFFY;

    return (user * HZ * CREDITS_PER_JIFFY) / XT_LIMIT_SCALE;
}
  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10

limit匹配处理

信用值使用32位存储,最大值0xFFFFFFFF。如下计算每个jiffy所对应的信用量值,内核将标准定义为最低速率一天一次。此时每个jiffy对应的信用值为(MAX_CPJ),满足最低速率要求。

#define MAX_CPJ (0xFFFFFFFF / (HZ*60*60*24))
  • 1

为了函数中的乘法能够被GCC在编译的时候使用左移实现,宏POW2_BELOW32将MAX_CPJ转换为2的幂值。首先宏_POW2_BELOW32将参数值x,通过右移位(>>)和或(|)操作将其有效位都设置为1(最高不为零的位到最低位都设置为1);之后,将结果右移移位,再加1。此值即为小于MAX_CPJ的最大的2的幂值,作为每个jiffy的信用值。

/* Repeated shift and or gives us all 1s, final shift and add 1 gives
 * us the power of 2 below the theoretical max, so GCC simply does a
 * shift. */
#define _POW2_BELOW2(x) ((x)|((x)>>1))                               b1000000000 | b100000000
#define _POW2_BELOW4(x) (_POW2_BELOW2(x)|_POW2_BELOW2((x)>>2))       b1100000000 | b11000000
#define _POW2_BELOW8(x) (_POW2_BELOW4(x)|_POW2_BELOW4((x)>>4))       b1111000000 | b111100
#define _POW2_BELOW16(x) (_POW2_BELOW8(x)|_POW2_BELOW8((x)>>8))      b1111111100 | b11
#define _POW2_BELOW32(x) (_POW2_BELOW16(x)|_POW2_BELOW16((x)>>16))   b1111111111 | b0
#define POW2_BELOW32(x) ((_POW2_BELOW32(x)>>1) + 1)                  b1111111111 >> 1 + 1

#define CREDITS_PER_JIFFY POW2_BELOW32(MAX_CPJ)
  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11

如下匹配函数limit_mt,信用值credit按照时间进行回填,根据前一次时间信用计算的时间(prev)与当前时间的jiffies差值,增加新的信用值,如果新的信用值大于最大值(credit_cap),重新设定为credit_cap的值。如果当前的信用值大于消耗(cost - 每个报文的消耗),表明还有余量,信用值减去消耗,返回真。否则,返回false,当前的报文速率过快,信用值不足。

static bool
limit_mt(const struct sk_buff *skb, struct xt_action_param *par)
{
    const struct xt_rateinfo *r = par->matchinfo;
    struct xt_limit_priv *priv = r->master;
    unsigned long now = jiffies;

    spin_lock_bh(&priv->lock);
    priv->credit += (now - xchg(&priv->prev, now)) * CREDITS_PER_JIFFY;
    if (priv->credit > r->credit_cap)
        priv->credit = r->credit_cap;

    if (priv->credit >= r->cost) {
        /* We're not limited. */
        priv->credit -= r->cost;
        spin_unlock_bh(&priv->lock);
        return true;
    }
    spin_unlock_bh(&priv->lock);
    return false;
  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20

iptables应用层处理limit

速率值在下发到内核前,扩大了XT_LIMIT_SCALE倍数。

static int parse_rate(const char *rate, uint32_t *val)
{   
    const char *delim;
    uint32_t mult = 1;  /* Seconds by default. */
    
    delim = strchr(rate, '/');
    if (delim) {
        if (strlen(delim+1) == 0)
            return 0;
        
        if (strncasecmp(delim+1, "second", strlen(delim+1)) == 0)
            mult = 1;
        else if (strncasecmp(delim+1, "minute", strlen(delim+1)) == 0)
            mult = 60;
        else if (strncasecmp(delim+1, "hour", strlen(delim+1)) == 0)
            mult = 60*60;
        else if (strncasecmp(delim+1, "day", strlen(delim+1)) == 0)
            mult = 24*60*60;
        else
            return 0;
    } 
    r = atoi(rate);
    if (!r) return 0;
    
    *val = XT_LIMIT_SCALE * mult / r;
  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25

内核版本 5.10

声明:本文内容由网友自发贡献,不代表【wpsshop博客】立场,版权归原作者所有,本站不承担相应法律责任。如您发现有侵权的内容,请联系我们。转载请注明出处:https://www.wpsshop.cn/w/Cpp五条/article/detail/632048
推荐阅读
相关标签
  

闽ICP备14008679号