关于 2019 年 4 月 2 日 XBTU19 和 ETHM19 的自动减仓事件

在北京时间 2019 年 4 月 2 日中午 12:44:34 与 13:22:08 之间,由于 XBTU19 和 ETHM19 的标记价格大幅波动,有不到 200 个仓位曾被自动减仓

在这些自动减仓事件发生时,分配给这些合约的保险基金是很少的。保险基金是根据特定合约(系统收益和损失)的强制平仓量分别分配给每份合约。对于已到期的合约,BitMEX 有一个流程将分配给这些合约的保险基金转入下一个即月合约。随着最近一次的到期是在 2019 年 3 月 29 日,这一过程失败,使即月合约没有得到重新分配,该资金维持未分配状态。因此,当这些受影响的合约发生大量的强制平仓事件时,少数用户曾被自动减仓。

BitMEX 收到自动减仓报告,并意识到异常高的自动减仓事件率,此时我们调查了此事。我们确定了根本原因,更正了其分配额度并设立了进一步的管控措施,以确保重新分配失败时会在内部发出警报。

对于受影响的用户,BitMEX 将亲自与您联系以解释情况以及您的补偿。我们根据用户在这些自动减仓事件的时间段内可能获得的最大潜在利润来补偿用户。我们以每份合约的最优惠价格为这些用户平仓:XBTU19 多头的价格为 5,079.5,ETHM19 空头的价格为 0.03103。 BitMEX 没有从这些自动减仓的仓位中获利。

对于给您带来的不便,我们深表歉意。 如果您有任何疑问,请联系客服。

欢迎转载,请注明文章来自

BitMEX (www.bitmex.com)

XBTU19 and ETHM19 Auto-Deleveraging Events 2 April 2019

On 2 April 2019 between 04:44:34 UTC and 05:22:08 UTC, less than 200 positions were auto-deleveraged due to the sharp price movements of the underlying mark price on XBTU19 and ETHM19.

At the time of these auto-deleveraging events, the Insurance Fund allocated to these contracts was minimal. The Insurance Fund is allocated individually to each contract according to how many liquidations contribute to that specific contract (System Gains and Losses). In the case of expired contracts, BitMEX has a process in place to roll over the Insurance Fund allocated to these contracts into the next front month contract. With the recent expiry on the 29th March 2019, this process failed and front month contracts did not receive their reallocation, and the funds remained unallocated. As a result, a handful of users were auto-deleveraged upon large liquidations within these affected contracts.

BitMEX receives auto-deleveraging reports and was made aware of the unusually high rate of auto-deleveraging events, at which point we investigated the matter. We identified the root cause, corrected the allocation and put further controls in place to ensure that reallocation failures are automatically flagged internally.

For users that were affected, BitMEX will be reaching out to you personally to explain the situation and document your compensation. We compensated users based on the maximum potential profit that they would have made over the timeframe of these auto-deleveraging events. We exited these users at the best price of each contract: longs at 5,079.5 on XBTU19 and shorts at 0.03103 on ETHM19. BitMEX did not profit from these auto-deleveraged positions.

We apologise for any inconvenience this caused. Should you have any questions, please contact customer support.

关于 2019 年 3 月 29 日服务轻微受阻的公告

在北京时间 3 月 29 号晚上 20:00,BitMEX 的服务曾轻微受阻约 15 秒,当时因正在进行结算,交易引擎处于封锁状态,所有请求已被卸载。受阻约 15 秒后,平台恢复正常。

在翌日凌晨 04:13,BitMEX 网站曾局部性中断服务,有一小部分的客户因此受到影响。问题已即时被发现和解决。API 服务没有受到影响。  

对您造成的不便,我们深表歉意。如果您有任何问题,请联系客服。

欢迎转载,请注明文章来自

BitMEX (www.bitmex.com)

Update: Notice of Minor System Outages 29 March 2019

On 29 March at 12:00 UTC, BitMEX experienced a minor outage for approximately 15 seconds whereby all requests would have been load shed as the engine was blocked during settlement operations. The platform was back to normal after the 15 second outage.

At 20:13 UTC, the BitMEX website experienced a limited interruption in service to a small group of users. The issue was immediately identified and fixed. The API was not affected.

We apologise for the inconvenience. Should you have any questions, please contact customer support.

关于 2019 年 2 月 19 日系统服务轻微受阻的公告

在北京时间 2 月 19 日下午 13:31 ,BitMEX 曾经历了大约 1 分钟的服务轻微受阻,导致所有交易引擎运作一度暂停。

问题是出于内部市场数据分发组件之间数据传输的持续时间。这曾经是定期更新的一部分,旨在提高平台的整体韧性。 我们已确定根本原因并通过内部流程进行修复以防止事件再次发生。 此外,我们正在继续重新设计我们的市场数据分布架构,以消除在这种情况下对交易引擎的任何潜在影响。

对于给您带来的不便,我们深表歉意。 如果您有任何疑问,请联系客服。

欢迎转载,请注明文章来自

BitMEX (www.bitmex.com)

Notice of Minor System Outage 19 February 2019

On 19 February at 05:31 UTC, BitMEX experienced a minor outage for approximately 1 minute whereby all trading engine operations were suspended.

This issue occurred due to a sustained period of data transfers between the internal market data distribution components. This was part of a regularly scheduled update to improve the overall resiliency of the platform. The root cause has been identified and a fix via internal processes has been put in place to prevent a recurrence. Additionally, we are continuing to re-work our market data distribution architecture to eliminate any potential impact to the trading engine in such a scenario.

We apologise for the inconvenience. Should you have any questions, please contact customer support.

有关 2019 年 2 月 8 日 API 超时的公告

今天在北京时间下午 13:40 和15:11 之间,由于 API 层的资源争用,有部分对 BitMEX REST API 的请求经历了缓慢的 API 响应, 最终导致 API 超时。通过我们的内部警报机制检测后,我们确定了原因,并在几分钟内减轻了直接影响。 目前没有持续的问题,在此期间对交易引擎或用户数据没有影响。
我们已经确定了针对该问题的根本原因的修复,并且正在将其作为优先事项进行处理。 一旦修复生效,我们将发布另一个公告跟进。 我们还提高了系统监控的灵敏度,以便更快地检测和解决潜在的类似问题。 对由此造成的任何不便,我们深表歉意。

欢迎转载,请注明文章来自

BitMEX (www.bitmex.com)

Notice of API Timeouts 8 February 2019

Between 05:40 and 07:11 UTC today, a subset of the requests to the BitMEX REST API experienced slow API responses and eventual API timeouts due to resource contention at the API layer. Upon detection via our internal alerting mechanisms we identified the cause and mitigated the immediate impact within a few minutes. There is currently no ongoing issue and there was no impact to the trading engine or user data during this time.

Fixes for the underlying root cause of the issue have been identified and are being worked on as a priority. We will follow up with another announcement once these are live. We have also increased the sensitivity of our system monitoring to detect and resolve potential similar issues much sooner. We apologise for any inconvenience this may have caused.

更新:解决上周问题的根由

针对上周的问题发布,昨天我们成功发布了内部市场数据分布组件的重新订阅逻辑的改进版。这解决了上周问题的根由和设立了额外的安全机制,以防止对当时所部署的交易引擎造成影响,我们预计上周的问题不会再次发生。

 

 

欢迎转载,请注明文章来自

BitMEX (www.bitmex.com)

Update: fix for root cause of last week’s issue

In response to last week’s post, yesterday we successfully released an enhancement to our internal market data distribution component’s re-subscription logic. This addresses the root cause of the previous week’s issue and along with the additional safety mechanism to prevent impact to the trading engine deployed at the time, we don’t anticipate a reoccurrence of last week’s issue.