← → or space to move between slides · f full screen · g all slides
Bi-weekly service review
Prepared for Tata Communications by Zixi
Live · last 14 days
4d 14h 46m of combined downtime across 69 outages, affecting 24 customers (1d 15h 34m elapsed). 5 unplanned incidents, the most serious a Sev 1 (full outage). 2 maintenance windows, 1 of them emergency. 2 incidents still open.
This period at a glance
All times UTC
Overview
Combined downtime
4d 14h 46m
across 69 outages · 1d 15h 34m elapsed
Longest single outage
21h 8m
Anuvu
ANUVU-BX-GPU-WEST-01
5
Unplanned incidents
2
Still open
24
Customers affected
2
Maintenance windows
1 emergency
Overview
Overview
Service impact
All times UTC.
Each row is one broadcaster, and each bar shows how long an outage lasted. Bars that line up across rows come from the same event. Showing the 8 broadcasters with the most downtime, of 57 affected; the full timeline is on the dashboard.
Service impact
Combined downtime by customer, region and host server.
Incidents
Incident 1 of 5
1d 17h 58m
Combined downtime
21h 8m
Elapsed time
3 Nov 19:47
Started (UTC)
2
Broadcasters affected
Customers affected: Anuvu
Both of Anuvu's Los Angeles broadcasters, on the same Tata KVM host, went down at 19:47 UTC on 3 November when the host failed. They came back at 16:38 and 16:56 UTC on 4 November, once the VMs had been rebuilt on another host and redeployed.
Root cause: A disk failure on Tata KVM host LA SR22 stopped the host from booting, taking down both VMs on it (Tata incident report TT 78004).
Resolution: Tata rebuilt both VMs on host LA SR07 by 13:30 UTC on 4 November; the broadcasters were redeployed on them and back by 16:56 UTC.
Incident 2 of 5
1d 2h 30m
Combined downtime
8h 51m
Elapsed time
17 Nov 05:10
Started (UTC)
3
Broadcasters affected
Customers affected: CS Test · Mercury Radio Arts (Glenn Beck) · Multimedios
Three Zixi VMs on Tata Cloudlyte in New Jersey went down from 05:10 UTC on 17 November. Tata's team is working to restore them.
Incident 3 of 5
19h 57m
Combined downtime
36m
Elapsed time
15 Nov 23:55
Started (UTC)
36
Broadcasters affected
Customers affected: Anuvu · Blaze TV · Cannella Media · Cinedigm · Estrella TV / Liberman Broadcasting · Family Broadcasting Corp. · Fox Networks · Impact Network Inc · Mercury Radio Arts (Glenn Beck) · Multimedios · National Black Television · National Black Television · RCN Television · SKY LINK TV · TCN The Country Network · TV5 USA · The Word Network · Tigertech Media · United Teleports · YouTubeTV [Production] · Zixi ContentXchange - CXE
Every Zixi VM in Los Angeles, on both Tata KVM and Cloudlyte, became unreachable from 23:55 UTC on 15 November until 00:31 UTC: 36 broadcasters across 20 customers.
Root cause: Urgent maintenance by Tata's backbone network team interrupted connectivity to the Los Angeles site (Tata incident report TT 78554).
Resolution: Service returned when the maintenance finished. Zixi confirmed every VM back by 00:31 UTC.
Incident 4 of 5
15h 57m
Combined downtime
5h 20m
Elapsed time
13 Nov 21:40
Started (UTC)
3
Broadcasters affected
Customers affected: Blaze TV · Zixi ContentXchange - CXE
Two Content Exchange ingest broadcasters and Blaze TV's broadcaster in New York were unreachable from 21:40 UTC on 13 November until 03:00 UTC on 14 November. Zixi classed it as a loss of redundancy.
Root cause: A fault on Tata's New York router. Tata's backbone network team reloaded the router to restore service.
Resolution: Router reloaded; all three broadcasters back by 03:00 UTC on 14 November.
Incident 5 of 5
0m
Combined downtime
0m
Elapsed time
4 Nov 19:35
Started (UTC)
—
Broadcasters affected
Since 4 November Charter has reported continuity errors on feeds it takes from Zixi's Los Angeles VMs over the Tata cross connect on VLAN 1732, among them FETV, Canal Sur and The Word Network. The feeds leave Zixi's VMs clean, and Charter's New York copy of FETV is clean.
Root cause: Loss is measured at one point only: 73,652 transmit discards on Tata's switch port toward Charter. Charter's port shows no input discards and no pause frames.
Resolution: Tata is moving the circuit to a switch from a different vendor; the parts were due on 18 November.
Maintenance
2 windows in the period, 1 of them emergency.
Action items
8 open of 8 tracked.
Trends
Completed reviews, each measure on its own scale.
Before the next review
Thank you. Review period 4–17 Nov 2026 · all times UTC