View Relay "cryShell02"
Fingerprint: 09587B9DE31CAAE07CDCDDB203D3D0153FE02D73 |
Family: 2 relays
|
88EACE21E04C30749711DBFF2CF89265F691B... |
AS13045
|
Germany
|
Linux
- Consensus
- In Consensus (9/9 Directory Auths)
- BW Verified
- 5/7 BW Auths Measured 35 Mbit/s Median
- Stability
- Overloaded | UP 10h 9m 55s ago | UP 96% (1M)
- Reachability
- IPv4: 9/9 | v6: None configured(Directory Auths)
- BW Weight
- 0.00% of Network | 39 Mbit/s Observed By Relay
- Version
- 0.4.9.15 Recommended
Issues Detected:
-
Guard: requires Stable flag
: Guard flag requires having the Stable flag first
Suggestion: Get Stable flag by maintaining consistent uptime. Stable requires uptime and MTBF at or above network median (typically 2-3 weeks of stable running). -
Guard: WFU below threshold
: 0/9 authorities meet guard-wfu (≥98%; dizum: ≥97.84099999999999%; 5 needed). bastet, dannenberg, dizum, faravahar, gabelmoo, longclaw, maatuska, moria1 and tor26 measured 86.2%–89.3%.
Suggestion: WFU measures recent uptime (recent downtime weighs more heavily). To improve: 1) Minimize restarts, 2) Use systemd with Restart=always, 3) Monitor for OOM kills, 4) Schedule updates during low-traffic periods. -
Not eligible for Stable flag
: 1/9 authorities meet Stable (uptime or MTBF at or above their own median; 5 needed). bastet, dannenberg, dizum, faravahar, gabelmoo, longclaw, maatuska and tor26 do not.
Suggestion: Stable flag requires uptime/MTBF at or above network median. Keep your relay running continuously for 2-3 weeks. Avoid restarts. Use reliable hardware and network connection. -
HSDir: requires Stable flag
: HSDir flag requires having the Stable flag first
Suggestion: Get Stable flag by maintaining consistent uptime. Stable requires uptime and MTBF at or above network median (typically 2-3 weeks of stable running). Avoid restarts. -
General Overload Active
: Relay reported general overload at 2026-11-04 03:00 UTC. Reported timestamps use hour precision, so the displayed minute is not an exact incident time. The flag does not identify which trigger fired. Investigation paths are Tor's internal OOM handling, dropped ntor onionskins, and TCP port exhaustion.
Suggestion: Inspect CPU and memory with htop, Tor service logs, host or container memory headroom, and the effective MaxMemInQueues. When the process was killed or restarted, check kernel OOM logs separately. A kernel OOM kill is a different mechanism from Tor freeing memory itself. MetricsPort: tor_relay_load_oom_bytes_total. Broken down by subsys. An increase measures bytes freed by Tor’s internal OOM handling. It is not an invocation count and it is not proof of a kernel OOM kill. Review Tor’s onionskin and drop messages, and whether the CPU is saturated or contended. The general-overload flag does not say this trigger fired. MetricsPort: tor_relay_load_onionskins_total{type="ntor",action="dropped"} and {type="ntor_v3",action="dropped"}. Compare each with the matching action="processed" series. Inspect both supported handshake types. Review Tor logs, socket usage, and the host’s ephemeral-port allocation. On Linux, ss -s and reading net.ipv4.ip_local_port_range are useful starting checks. MetricsPort: tor_relay_load_tcp_exhaustion_total. An increase records additional exhaustion events. See the Tor overload troubleshooting guide.
Addresses
- OR Address
-
a89-182-151-21.net-htp.de
89.182.151.21:9001
Reachability (Directory Authorities)
- IPv4
- 9/9
- IPv6
- None configured
Location
- Country
-
Germany
- Interactive Map
- View in RouteFluxMap
Eligible Flags (4):
Running: 9/9 | Valid: 9/9 | V2Dir: 9/9 | Fast: 9/9 | Stable: 1/9 | HSDir: 0/9 | Guard: 0/9 | Exit: 0/9
Eligibility Flag Vote Details
| Flag | Metric | Status | Relay Value | Threshold Required |
|---|---|---|---|---|
| Running | IPv4 Reachability | Meets (9/9 DA) | 9/9 authorities reached relay | ≥5/9 DA ORPort reachable |
| Valid | Descriptor | Meets (9/9 DA) | 9/9 authorities assigned Valid | Not blacklisted + valid descriptor (≥5/9 DA) |
| V2Dir | DirPort Available | Meets (9/9 DA) | Tunnelled: Yes (no DirPort) (R) | Tunnelled directory via ORPort or DirPort (≥5/9 DA) |
| Fast | Speed | Meets (9/9 DA) | 32.8 Mbit/s – 41.6 Mbit/s (5 DA measured) 39.8 Mbit/s (4 DA relay-reported) |
≥816.0 Kbit/s (AuthDirFastGuarantee, or the top-7/8 cutoff if lower) (≥5/9 DA) [Stricter] moria1: ≥8.4 Mbit/s |
| Stable | MTBF | Below (1/9 DA) | Median: 0.7d (1/9 DA above threshold) Min/Max: 0.5d / 19047.5d (9/9 DA, Voting / Total Eligible) |
≥2.4d - 9.3d (varies) (≥5/9 DA) [Stricter] moria1: ≥7607.5d |
| Uptime | Below (0/9 DA) | 0.4d (R) | ≥5.2d - 5.3d (varies) (≥5/9 DA) | |
| HSDir | Prereq: Fast | Meets (9/9 DA) | 9/9 DA assigned Fast 32.8 Mbit/s – 41.6 Mbit/s (5 DA measured) 39.8 Mbit/s (4 DA relay-reported) |
≥816.0 Kbit/s (AuthDirFastGuarantee, or the top-7/8 cutoff if lower) (≥5/9 DA) [Stricter] moria1: ≥8.4 Mbit/s |
| Prereq: Stable | Below (1/9 DA) | 1/9 DA assigned Stable | ≥2.4d - 9.3d (varies) (≥5/9 DA) [Stricter] moria1: ≥7607.5d |
|
| Prereq: V2Dir | Meets (9/9 DA) | 9/9 DA assigned V2Dir | Tunnelled directory via ORPort or DirPort (≥5/9 DA) | |
| Uptime | Below (0/9 DA) | 0.4d (R) | ≥96h as each authority sees it (MinUptimeHidServDirectoryV2) (≥5/9 DA) [Stricter] moria1: ≥98% WFU and ≥9.8d Time Known |
|
| Guard | Prereq: Fast | Meets (9/9 DA) | 9/9 DA assigned Fast 32.8 Mbit/s – 41.6 Mbit/s (5 DA measured) 39.8 Mbit/s (4 DA relay-reported) |
≥816.0 Kbit/s (AuthDirFastGuarantee, or the top-7/8 cutoff if lower) (≥5/9 DA) [Stricter] moria1: ≥8.4 Mbit/s |
| Prereq: Stable | Below (1/9 DA) | 1/9 DA assigned Stable | ≥2.4d - 9.3d (varies) (≥5/9 DA) [Stricter] moria1: ≥7607.5d |
|
| Prereq: V2Dir | Meets (9/9 DA) | 9/9 DA assigned V2Dir | Tunnelled directory via ORPort or DirPort (≥5/9 DA) | |
| WFU | Below (0/9 DA) | Median: 88.5% (0/9 DA above threshold) Min/Max: 86.2% / 89.3% (9/9 DA, Voting / Total Eligible) |
≥98%; dizum: ≥97.84099999999999% (≥5/9 DA) | |
| Time Known | Meets (9/9 DA) | Median: 10.3d (9/9 DA above threshold) Min/Max: 9.8d / 10.6d (9/9 DA, Voting / Total Eligible) |
≥8 days (≥5/9 DA) | |
| Bandwidth | Meets (9/9 DA) | 32.8 Mbit/s – 41.6 Mbit/s (5 DA measured) 39.8 Mbit/s (4 DA relay-reported) |
≥16.8 Mbit/s (AuthDirGuardBWGuarantee) OR top 25% (≥80.0 Mbit/s-272.0 Mbit/s) (≥5/9 DA) | |
| Exit | Exit Policy | Below (0/9 DA) | Port 80: No | Port 443: No (R) | Allows ≥1 /8 on ports 80 AND 443 (≥5/9 DA) |
(DA) = Dir. Auth. Measured |
(R) = Relay Reported
Data from Tor CollecTor (authority votes, fetched 2026-11-04 12:45:10).
Capacity (Relay Reported)
- Observed Bandwidth
- 39.82 Mbit/s
- Advertised Bandwidth
- 39.82 Mbit/s
- Rate Limit
- 41.94 Mbit/s
- Burst Limit
- 67.11 Mbit/s
- Total Data
-
56 TB
(5 years)
1mo: 4.8 TB | 6mo: 24 TB | 1y: 40 TB | 5y: 56 TB
Measurement (Directory Authority Verified)
Source: CollecTor — fetched 2026-11-04 12:45:10
- Measured By
-
5/7
Bandwidth Authorities
Not measured by: faravahar, longclaw - Consensus Weight
- 35.2 Mbit/s (median of 5 measurements)
- Measured Min / Max
- 32.8 Mbit/s / 41.6 Mbit/s
Network Participation:
Consensus Weight:
0.0022% of network
(30th pctl)
|
Guard:
N/A |
Middle:
0.0063% (57th pctl) |
Exit:
N/A
History
Uptime Metrics
- Current Status
- UP 10h 9m 55s ago
- Overall Uptime (1M/6M/1Y/5Y)
- 96.8%/96.6%/96.7%/97.3%
- Flag Uptime (1M/6M/1Y/5Y)
- Matches Overall Uptime (96.8%/96.6%/96.7%/97.3%)
- Hibernating
- No
Timestamps
- First Seen
- 1y 4mo 2w ago (2025-06-20 10:00:00)
- Last Seen
- 2h 45m 39s ago (2026-11-04 11:00:00)
- Last Restarted
- 10h 9m 55s ago (2026-11-04 03:35:44)
- Last Changed Address
- 2026-11-04 05:00:00
[Warning] OVERLOADED
— General overload at 2026-11-04 03:00 UTC
Evaluated 2026-11-04 13:45 UTC. /details snapshot 2026-11-04 11:00:00 UTC (cached 42s ago). /bandwidth snapshot 2026-11-04 05:00:00 UTC (cached 6h ago).
Reported timestamps use hour precision, so the displayed minute is not an exact incident time.
General Overload (from /details)
- Status
- Reported
- When
- 2026-11-04 03:00 UTC (10 hours ago)
The general-overload flag does not say which trigger fired. Possible paths are Tor’s internal memory handling, dropped ntor handshakes, and TCP port exhaustion. Allium treats a valid report as eligible for 72 hours; that cutoff is not proof the relay recovered then.
File Descriptor Exhaustion (from /bandwidth)
- Status
- No file-descriptor exhaustion report in available data
Rate Limits (from /bandwidth)
- Status
- No rate-limit report in available data
- Rate Limit
- —
- Burst Limit
- —
- Write counter
- —
- Read counter
- —
What to check
-
Tor internal memory pressure
Inspect CPU and memory with htop, Tor service logs, host or container memory headroom, and the effective MaxMemInQueues. When the process was killed or restarted, check kernel OOM logs separately. A kernel OOM kill is a different mechanism from Tor freeing memory itself.MetricsPort:
tor_relay_load_oom_bytes_total— Broken down by subsys. An increase measures bytes freed by Tor’s internal OOM handling. It is not an invocation count and it is not proof of a kernel OOM kill. -
ntor handshake pressure
Review Tor’s onionskin and drop messages, and whether the CPU is saturated or contended. The general-overload flag does not say this trigger fired.MetricsPort:
tor_relay_load_onionskins_total{type="ntor",action="dropped"} and {type="ntor_v3",action="dropped"}— Compare each with the matching action="processed" series. Inspect both supported handshake types. -
TCP port exhaustion
Review Tor logs, socket usage, and the host’s ephemeral-port allocation. On Linux, ss -s and reading net.ipv4.ip_local_port_range are useful starting checks.MetricsPort:
tor_relay_load_tcp_exhaustion_total— An increase records additional exhaustion events.
About these reports and MetricsPort
Allium’s eligibility windows are applied to the reported timestamps: 72 hours for general overload and file-descriptor exhaustion, 24 hours for a rate-limit hit. A report at the cutoff is historical. An older report can remain visible after Onionoo or Allium’s cache still has it. The cutoff does not prove the relay recovered at that moment. Allium caches the bandwidth document for 12 hours by default. The snapshot time is when Onionoo published that document. It is not the relay’s event time, and it is not necessarily the extra-info publication time. Cache age is how old Allium’s copy is.
MetricsPort counters are cumulative. Compare samples, and account for process restarts. A nonzero total alone does not show ongoing activity. Gauges describe current values. An unavailable metric is unknown, not zero. Allium does not have this relay’s MetricsPort readings and does not invent them.
The port below is an example for an endpoint you have already configured. MetricsPort must be enabled with restricted access and must not be exposed publicly. Setup stays in the Tor overload troubleshooting guide.
curl http://127.0.0.1:9035/metrics
History
Operator Identity
Source: AROI Validation API — 2026-11-04 04:00 UTC
- AROI Domain (?)
- Not specified
- Operator Validation
- Unvalidated
- This Relay
- N/A (no AROI configured)
- Contact
- 88EACE21E04C30749711DBFF2CF89265F691B280 cryShell <kx@cryshell.com>
Family Configuration
- Family Support
- Happy Families (2) — Has family-cert in server descriptor (guide) | View Full Family (2)
Happy Families (family-cert)
- Effective Members: 2
09587B9DE31CAAE07CDCDDB203D3D0153FE02D73 (this relay) CD1D703E667366A5474E4FF6B1BE6D0C9178B2B0
MyFamily (legacy declarations) — none
DNS Health
Not an exit relay — DNS health testing does not apply.Last fetch was at Wed, 04 Nov 2026 13:45:39 GMT.
Data from Tor CollecTor (authority votes, fetched 2026-11-04 12:45:10).
Each cell shows: your measured value | authority threshold. Green = meets, red = below. Flags: green = all authorities agree, yellow = partial.
| Authority | Running | Valid | Exit | MiddleOnly | BW Scan (Authority) |
v4 | Flags | Fast (BW|T) | Guard BW (BW|Threshold) | Guard WFU (M|T) | Guard TK (M|T) | Stable MTBF (M|T) | Stable Uptime (Relay|T) | Desc Published | Cons Wt (Authority) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| bastet (US) | Yes | Yes | No | — | Y | Yes | Running, Valid, V2Dir, Fast | 40.0 Mbit/s (M) | 816.0 Kbit/s | 40.0 Mbit/s (M) | 192.0 Mbit/s OR 16.8 Mbit/s | 89.3% | 98% | 10.4d | 8d | 0.7d | 8.4d | 0.4d | 5.2d | 2026-11-04 04:05:15 | 40.0 Mbit/s (M) |
| dannenberg (DE) | Yes | Yes | No | — | Y | Yes | Running, Valid, V2Dir, Fast | 32.8 Mbit/s (M) | 816.0 Kbit/s | 32.8 Mbit/s (M) | 272.0 Mbit/s OR 16.8 Mbit/s | 88.6% | 98% | 10.6d | 8d | 0.7d | 9.3d | 0.4d | 5.3d | 2026-11-04 04:05:15 | 32.8 Mbit/s (M) |
| dizum (NL) | Yes | Yes | No | — | N | Yes | Running, Valid, V2Dir, Fast | 39.8 Mbit/s (R) | 816.0 Kbit/s | 39.8 Mbit/s (R) | 80.0 Mbit/s OR 16.8 Mbit/s | 86.2% | 98% | 10.1d | 8d | 0.7d | 6.7d | 0.4d | 5.3d | 2026-11-04 04:05:15 | 39.8 Mbit/s (R) |
| faravahar (US) | Yes | Yes | No | — | Y | Yes | Running, Valid, V2Dir, Fast | 39.8 Mbit/s (R) | 816.0 Kbit/s | 39.8 Mbit/s (R) | 80.0 Mbit/s OR 16.8 Mbit/s | 88.7% | 98% | 10.5d | 8d | 0.7d | 6.5d | 0.4d | 5.3d | 2026-11-04 04:05:15 | 39.8 Mbit/s (R) |
| gabelmoo (DE) | Yes | Yes | No | — | Y | Yes | Running, Valid, V2Dir, Fast | 41.6 Mbit/s (M) | 816.0 Kbit/s | 41.6 Mbit/s (M) | 248.0 Mbit/s OR 16.8 Mbit/s | 88.3% | 98% | 10.1d | 8d | 0.7d | 10.8d | 0.4d | 5.3d | 2026-11-04 04:05:15 | 41.6 Mbit/s (M) |
| longclaw (US) | Yes | Yes | No | — | Y | Yes | Running, Valid, V2Dir, Fast | 39.8 Mbit/s (R) | 816.0 Kbit/s | 39.8 Mbit/s (R) | 80.0 Mbit/s OR 16.8 Mbit/s | 88.5% | 98% | 10.3d | 8d | 0.7d | 10.9d | 0.4d | 5.3d | 2026-11-04 04:05:15 | 39.8 Mbit/s (R) |
| maatuska (SE) | Yes | Yes | No | — | N | Yes | Running, Valid, V2Dir, Fast | 39.8 Mbit/s (R) | 816.0 Kbit/s | 39.8 Mbit/s (R) | 80.0 Mbit/s OR 16.8 Mbit/s | 87.2% | 98% | 10.0d | 8d | 0.5d | 2.4d | 0.4d | 5.3d | 2026-11-04 04:05:15 | 39.8 Mbit/s (R) |
| moria1 (US) | Yes | Yes | No | — | Y | Yes | Running, Valid, V2Dir, Fast, Stable | 34.4 Mbit/s (M) | 8.4 Mbit/s | 34.4 Mbit/s (M) | 192.0 Mbit/s OR 16.8 Mbit/s | 88.3% | 98% | 9.8d | 8d | 19047.5d | 7607.5d | 0.4d | 5.3d | 2026-11-04 04:05:15 | 34.4 Mbit/s (M) |
| tor26 (AT) | Yes | Yes | No | — | Y | Yes | Running, Valid, V2Dir, Fast | 35.2 Mbit/s (M) | 816.0 Kbit/s | 35.2 Mbit/s (M) | 256.0 Mbit/s OR 16.8 Mbit/s | 88.6% | 98% | 10.4d | 8d | 0.7d | 10.6d | 0.4d | 5.3d | 2026-11-04 04:05:15 | 35.2 Mbit/s (M) |
N/T = this authority does not test IPv6 • — = not in this vote • Format: relay value | authority threshold • WFU = Weighted Fractional Uptime • TK = Time Known • Bandwidth source: (M) bandwidth-scanner measurement, (R) relay-reported, (U) unmeasured, counted as 0
Bandwidth Values Explained:
Measured (M) = An authority's bandwidth-scanner measurement of your relay (
Relay-reported (R) = The bandwidth your relay advertises: the lower of its observed bandwidth and its rate limit, capped at 10 MB/s (
Not published = The authority has no measurement of your relay in its vote and ignores relay-reported bandwidth, yet it voted Fast or its vote has no measurements at all: it judges your relay by a measurement kept from an earlier bandwidth file (0 if none), which its vote doesn't show.
Why different? A low measurement means the bandwidth scanners see your relay underperforming compared with other relays that report a similar speed. Votes list both values in kilobytes per second. Rules: Assigning flags in a vote, Computing a consensus.
Measured (M) = An authority's bandwidth-scanner measurement of your relay (
Measured= in its vote). An authority that has one uses it for the Fast and Guard flags, and the consensus weight is the median of these measurements when 3 or more authorities have one.Relay-reported (R) = The bandwidth your relay advertises: the lower of its observed bandwidth and its rate limit, capped at 10 MB/s (
Bandwidth= in every vote). An authority without a measurement uses it for flags, unless it has measurements for enough other relays (ignoring-advertised-bws=1); then your relay counts as 0 (U). It sets the consensus weight only when fewer than 3 authorities measured the relay, capped at 20 KB/s (maxunmeasuredbw) while 3 or more authorities publish measurements.Not published = The authority has no measurement of your relay in its vote and ignores relay-reported bandwidth, yet it voted Fast or its vote has no measurements at all: it judges your relay by a measurement kept from an earlier bandwidth file (0 if none), which its vote doesn't show.
Why different? A low measurement means the bandwidth scanners see your relay underperforming compared with other relays that report a similar speed. Votes list both values in kilobytes per second. Rules: Assigning flags in a vote, Computing a consensus.
Stable Uptime (Two Data Sources):
Relay Uptime = From Onionoo API (last_restarted field). Self-reported uptime from descriptor. Same value for all authorities.
Authority Threshold = From CollecTor vote files (flag-thresholds stable-uptime). Each authority may have different thresholds.
Why two sources? Vote files contain authority thresholds but not per-relay uptime values. Onionoo aggregates relay descriptors which include uptime.
Relay Uptime = From Onionoo API (last_restarted field). Self-reported uptime from descriptor. Same value for all authorities.
Authority Threshold = From CollecTor vote files (flag-thresholds stable-uptime). Each authority may have different thresholds.
Why two sources? Vote files contain authority thresholds but not per-relay uptime values. Onionoo aggregates relay descriptors which include uptime.