Telstra service status: outage reports and connection issues

Problems in the last 24 hours

The graph below depicts the number of Telstra reports received over the last 24 hours by time of day. When the number of reports exceeds the baseline, represented by the red line, an outage is determined.

Most Reported Problems

The following are the most recent problems reported by Telstra users through our website.

• Phone (54%)
• Internet (28%)
• Total Blackout (8%)
• Wi-fi (5%)
• E-mail (4%)
• TV (1%)

Live Outage Map

The most recent Telstra outage reports came from the following cities: Sydney, Brisbane, Melbourne, Maleny, Perth, Nowra, Adelaide, Bunbury, Gold Coast, Wasleys, Townsville, Canberra, Geelong, Arumpo, and Gympie.

City	Problem Type	Report Time
Sydney	Phone	2 hours ago
Brisbane	Internet	15 hours ago
Melbourne	Internet	21 hours ago
Maleny	Internet	24 hours ago
Sydney	Internet	1 day ago
Brisbane	Phone	1 day ago

Full Outage Map

Community Discussion

Tips? Frustrations? Share them here. Useful comments include a description of the problem, city and postal code.

Beware of "support numbers" or "recovery" accounts that might be posted below. Make sure to report and downvote those comments. Avoid posting your personal information.

Telstra Issues Reports

Latest outage, problems and issue reports in social media:

• 
  Nathan ***** (@NathanOrgan) reported 8 minutes ago

  @grok @RositaDaz48 Telstra lmfao!! Charter Hall? lol Langdon capital? Are these people able to understand public service and the true value of work? I doubt it.

• 
  jordanjackelyse (@jordanjackelyse) reported 9 minutes ago

  @noahitallll @PatricPaal @pelisonlinegrat Work choices, joined the Iraq war, sold off last shares commonwealth bank, Telstra, and long term leases for airports, and then the biggest one of all. His 2002 China LNG (liquefied natural gas) agreement signed was the worst resource deal in Australian history. He was bad. 🤡

• 
  Michael (@MSVLKnight) reported 10 minutes ago

  @LisaLemon09 I worked for Telstra for a long time and in my experience nearly every major issue was caused by updates being tested in a lab with 10 services and then getting rolled out to 20,000 plus subscribers. It was nearly always inadequate testing.

• 
  bitterboy (@twistedandbitte) reported 10 minutes ago

  @Telstra What a bloody disgrace you are Telstra. Heard back from the person the complaint went to. Their response was instead of discussing the issue bluntly told us that given they have called us the case is now closed. Never discussed anything refused to discuss anything.

• 
  Smokey Dawson (@farcanal53) reported 12 minutes ago

  Whats wrong with our economic system is the unpenalised rank hypocrisy when Telstra declares a record profit, its CEO gets a $6 million payrise after the recent triple 0 outage that claimed lives.

• 
  🐏 (@princefishey) reported 12 minutes ago

  @aphexnaim CLASSIC VLINE last time i went to the city it was when the telstra outage messed all the trains up

• 
  leki ⚔️ (@mkfilko) reported 17 minutes ago

  $BRUN and $SHAZ are the two small cap neoclouds people keep putting in the same bucket. Somewhat similar market caps, both Nasdaq listed this year, both with an NVIDIA story attached. I think both provide compelling reasons to invest your money into, however, I do believe BRUN is the superior neocloud. I will present the facts as they are and sprinkle some commentary in it. You can use the facts to decide which is better for you. TL;DR at the bottom. Start with contract book versus actual revenue. Q2 2026, same quarter, both companies: > $BRUN revenue: $31.1M, up ~270% YoY > $SHAZ revenue: $1.9M, up ~412% YoY > $BRUN total contract value: $1.9B > $SHAZ total contract value: $8.8B $SHAZ carries 4.6 times the contract book on about one sixteenth of the revenue. Put differently, $BRUN's backlog is roughly 15 times its current annualized run rate. $SHAZ's is closer to 1,100 times. $SHAZ management guides first material revenue to Q4 2026. $BRUN has six sites in production today, took ARR from $30M at the end of 2025 to $145M by June, and reaffirmed roughly $400M exit ARR. A contract book is a claim while revenue is the confirmation of that claim. $BRUN has so far managed to prove that it is able to convert the contract book better than $SHAZ. Moving to the balance sheet, which is where $SHAZ shines > $SHAZ ended Q2 with roughly $1.9B in cash against $BRUN's $134.7M as of August 12. It has raised about $2.2B since December 2025, including a $1.6B oversubscribed round in June. On funding capacity this is not close, and $SHAZ wins it decisively. > The NVIDIA arrangement. Six years, $4.88B, up to 40,000 GB300 GPUs on NVIDIA's DSX AI factory design, 72MW in Australia, with NVIDIA taking product revenue plus a share of cloud revenue on supported capacity. Management describes the guaranteed pricing as a floor rather than a ceiling. $BRUN has no equivalent. This is the strongest single item in the $SHAZ file and anyone comparing the two has to concede it. Next, counterparty/customer quality: On April 1 $SHAZ announced a five year, $1.25B AI cloud infrastructure agreement, roughly 8,200 B300 GPUs deployed in Australia. The counterparty is ESDS Software Solution, an Indian cloud and managed services provider. ESDS is a real company. Five data centers in India, 2,501 customers as of FY26, growing revenue at a 28% CAGR. Using its own most recent disclosed accounts: > ESDS FY26 revenue: Rs 472 Cr, roughly $54M > ESDS FY26 profit after tax: Rs 121 Cr, roughly $14M > Annual obligation under the $SHAZ contract: ~$250M The yearly payment is about 4.7 times the counterparty's entire annual revenue and roughly 18 times its net profit. ESDS filed a Rs 720 Cr IPO in India on August 28, which raises roughly $82M against a five year obligation of $1.25B. Nothing here requires assuming bad faith by anyone. It requires asking how a company earning $14M a year funds $250M a year, and that question does not yet have a public answer. ESDS was not named on the $SHAZ Q2 call. When compared to $BRUN's own customer book: $BRUN's weaknesses are the concentration of their revenues. The top three customers were about 76% of 2025 revenue, with RunPod alone, a private GPU cloud platform, at about 45%. Both figures fell hard from 2024 on revenue that more than doubled, so concentration is getting lower. However, I think there is still room for improvement here. The quality side. The largest new contract signed is Thinking Machines Lab, roughly $470M over three years for about 5,000 GPUs, which implies around $3.60 per GPU hour. Mira Murati's lab raised $2B at a $12B valuation and has been reported in talks at a materially higher one since. That is a counterparty that can fund a $470M commitment out of capital already raised. Additionally there are the two agreements sitting in the Q2 10-Q, which are the part I find most useful because the cash terms are filed rather than announced: > June 8 agreement, 240 GPU servers, three year initial term, total consideration ~$207.6M. Structure: $18.97M prepaid before services commence, a further $37.94M on the commencement date, then ~$3.69M a month. That is about 27% of contract value collected at or before day one, and the first $18.97M was already sitting in customer deposits at June 30. > July 9 order form, 192 servers and 1,536 B300 GPUs, 48 month term, total contract value ~$222.5M. A prepayment equal to 25% of contract value, approximately $55.6M, not refundable, due in installments through service start. Two contracts, roughly $430M combined, each with a quarter of the money committed before a GPU is deliveredKaros on the call: > "We obtain prepayment on every deal we close." Average prepayment is 22% of TCV, and that single design choice runs a counterparty credit test automatically, at signing, on every contract. A $1.25B agreement under this structure would require roughly $275M wired before a single rack is energized. A counterparty that cannot post it never becomes backlog in the first place. Guckel confirmed what has actually been collected: > "we held $128.4 million in total customer deposits" $34.9M current, $93.5M long term, against compute scheduled for 2026 and 2027 delivery. So $BRUN's book is concentrated and its largest customer is a reseller, both fair criticisms. It is also a book where the customers have already wired a fifth of the money. Backlog funded in cash is a different asset class from backlog that has only been announced. Onto leadership, which is where most of my confidence actually comes from. On paper, $SHAZ has the more directly relevant development resume. James Manning has built and monetized more than 300MW of energy and compute infrastructure across Pennsylvania, Georgia, Texas and Australia, and founded a Nasdaq listed digital infrastructure company before this one. Nick Hughes-Jones ran 100 modular data centers across 200MW. The board is genuinely credentialed: Andrew Penn AO, former Telstra CEO, chairs it, and Western Union's chief legal officer sits on it. Anuj Goel joined as CFO from a 20 year Macquarie career running technology for APAC. Commercially they have people out of IBM, Microsoft, Rackspace and Equinix. This is not a shell. Three things still separate the two teams for me. > Tenure and continuity. $SHAZ's chief executive took the role in January 2026, its chairman arrived in May, its chief legal officer inside the last year, and its CFO in August. Board average tenure is roughly 0.6 years. That is a leadership team assembled over about twelve months, now asked to deliver 212MW. $BRUN's is the same group twice: Karos and Georgakopoulos both built Blue Fire Capital, both went to Galaxy Digital when it was acquired, and both left together to build this. Second company, same bench, and Georgakopoulos has been COO since April 2024. > They were the demanding customer before they were the vendor. Karos was MD and Head of Electronic Trading at Galaxy Digital and an executive committee member, and before that co founded and ran Blue Fire Capital across six countries and thirteen data centers. Karim Ali, the CIO, spent close to twenty years in performance sensitive infrastructure, built ultra low latency trading networks across four continents on microwave, millimeter wave and global fiber, and led a FedRAMP certification. In high frequency trading, latency and downtime hit the operator's own P&L the same day. This team ran compute as a cost center where failure was expensive to them personally, then went and sold it. That is a different instinct from developing a site and leasing it out. > Financing sits inside the company. Erik Guckel holds a PhD in Chemical Engineering from Illinois alongside a Chicago MBA, has closed over $2B in corporate transactions, managed a $3B debt portfolio, and secured funding for first of a kind facility construction. Karos has a derivatives and mathematical background and a track record monetizing billions in credit facilities. On the Q2 call he worked the prepayment against capex arithmetic live rather than passing it to the CFO, which is pretty damn impressive Both teams have now had one quarter to show what they do with a promise. $BRUN's first print as a public company came in at $31.1M with guidance reaffirmed. $SHAZ reported $1.93M against roughly $7.54M expected, a miss of about 75%, and moved first material revenue to Q4. I acknowledge that one quarter proves very little on its own. However, it is still the only head to head delivery test either team has actually run. Apart from the balance sheet mentioned earlier, $BRUN is also arguably weaker in its share structure and governance risks (dual-edged sword): Applying a governance screen to $SHAZ means applying the same screen to $BRUN. It has a dual class structure, roughly 50M Class A against 29.5M Class B. It had a related party loan from its own CEO, since repaid at closing. A director received a 336,000 share consulting grant that fully vested inside one quarter. And 7,875,000 CEO earnout shares plus sponsor and SPV earnouts all hit their price targets and vested within a single quarter of listing. These are ordinary **** era structures on both sides. Neither company is a governance exemplar What I am watching next: The NVIDIA financing gap is the strongest argument for $SHAZ, and $BRUN management spent a meaningful portion of the Q2 call signalling they are working on the same thing. Karos, on the AI Cloud Partner program: > "It's a well-known publicized program with NVIDIA, and obviously NVIDIA's investment grade. At the appropriate time, we can get into more details on that." Later, on the $500B consortium: > "And same with the AICP. We think that that program will continue to get extremely interesting, and I think we're going to have a plentiful amount of optionality currently and go forward." Guckel, separately: > "We continue to innovate on financing structures beyond the equipment financing approach used to date and expect to provide additional color and progress on this over the next quarter." And on the relationship itself: > "together, we are exploring deployments at scale that could meaningfully expand our footprint" None of that is signed. Management is definitely teasing a NVDA-linked financing structure during the call. But the specific advantage $SHAZ holds today is the specific advantage $BRUN is signalling it is negotiating, with a next quarter timeline attached to it, from a company NVIDIA already granted exemplar status and audits cluster by cluster. TL;DR: $SHAZ has the better balance sheet and the better NVIDIA arrangement, today, and neither of those is a small thing. $BRUN has sixteen times the revenue, a contract book that customers have already funded a fifth of in cash, six sites in production, and a counterparty filter built into how it signs deals. Between a large announced book and a smaller collected one, I think the collected one is a safer and yet equally asymmetric. Thanks for reading!

• 
  Ray G Kosanjic (@kosanjic) reported 19 minutes ago

  @PeterD84508 @DHughesy You idiot. Both sides of parliament agreed to sell. It was costing Australians hundreds of millions as a public service model. If Telstra was government owned now we’d be further down a hole.

• 
  Lucy Graham (@montrosegraham) reported 22 minutes ago

  @ChinaSelect Australia has had two major telecommunications network failures. As a result of the Telstra failure, three people died, including a baby. One person died as a result of the Optus failure. No infrastructure should have any exposure to foreign adversaries.

• 
  Rohanc (@rohancct) reported 22 minutes ago

  Forget the Telstra outage, there should be a senate investigation into AFL Tables being down for hours.

• 
  Stu (@StuartFromAus) reported 26 minutes ago

  @dan_const @AFL @Telstra William dropped off towards the end if the season. Also had one of his worst games in a rare prime time game in Vic, I knew it was done after that.

• 
  Kråke (@Kroke14) reported 33 minutes ago

  @7NewsSydney Great now she'll offshore even more Telstra workers under the guise of AI adoption so she can recoup some money for herself. Worst Telstra CEO ever..

• 
  Ivor B (@IvorB6) reported 33 minutes ago

  @danderfield 3/ one of the players who showed leadership following the Wayne Carey incident. Left his footy boots in the middle of Telstra Dome following Norths EF loss in 2005, notifying his retirement from the game.

• 
  Hugh_96 (@Hugh_96) reported 34 minutes ago

  @susanmandrews RA seems to be taking quite a unique approach to the women’s rugby program. We have all been through it. Unless it is a significant issue which requires an apology eg Optus/Telstra outages, just put out a holding statement & some spin. RA is lucky not much media interest in rugby

• 
  Roller D (@RollerD3) reported 35 minutes ago

  "Something went wrong" @Telstra The biggest POS company in Australia. Its Mind boggling how **** the sim card activation process it. ******* useless.

• 
  Howard (@howardw46) reported 40 minutes ago

  @maccaburbsvoice @Telstra This was about Telstra service ie phone connectivity!

• 
  Baiyang **** (@NJCMLAB) reported 45 minutes ago

  @MinotaurStocks In the Q2 conference call, it stated Telstra (Australia) as the new HomeSecure partner. "Third, we secured a new win ---, adding our HomeSecure service in another country. The HomeSecure solution enhances threat protection across the Telstra's ---"

• 
  DemocracyIsDead (@STIIBTWTCFMYCGA) reported 46 minutes ago

  @TopherField One of our clients had an existing telstra router just wanted an NBN connection. 2 weeks. $110 per month. I got her to contact starlink. 5 days later delivered and installed. $75 per month. Get your act together telstra. Can't be so slow anymore. There are choices.

• 
  vinni • 包沁燕 🇵🇸 (@glyphclutter) reported 46 minutes ago

  @simianlines i used to be with vodafone but now i’m on telstra wholesale network so what ******* gives

• 
  Dane Trethowan (@grtdane) reported 46 minutes ago

  @ellensandell Poor demented greens, Telstra, Optus etc have had Data Centers around the place for years.

• 
  Hung Long 🇱🇧🇵🇸🇮🇶🪃 (@Elmer_Hauser) reported 48 minutes ago

  Realistically, Trembath should finish 3rd in the Telstra Rising Star after Jagga and Dean, even just for consistency's sake. Wilma Durrsma's last 6 weeks has been utter garbage #AFLNorthCats

• 
  Lucy the Red (@LucytheRed1) reported 48 minutes ago

  @DHughesy You clearly don’t realize the Future Fund was: 1) primarily seeded by the sale of Telstra 2) Set up to fund existing government liabilities I’ll dumb it down for you. The Howard government sold our assets to fund existing liabilities. Embarrassed for you (again)

• 
  GibberishCapital (@GibberCapital) reported 48 minutes ago

  @ellensandell The data centres are built by Australian companies you utter imbecile. Do you have problem with data centres built by Goodman? NextDC? Used by Telstra? Optus? Vodafone? How are you such a ******* simpleton?

• 
  afraid_au (@afraid_au) reported 49 minutes ago

  @LmDread @tarkov Telstra appears to have stabilised for Sydney. Which servers are you having issues with, and which ISP are you with?

• 
  outsider (@outsider2040) reported 50 minutes ago

  @Urquie_Coffs @TonyShepherd4 Of course he left a measly surplus after selling off Qantas Telstra airports and many other government assets. Can be done once but never again so it pumped his numbers up big time. Other than that he pulled in some small tax receipts from massive mining boom.

• 
  BOB ACHMAR (@maroniteMAGA) reported 52 minutes ago

  @FreedmFightr1 Too many personal info,,, ffs Telstra origin and how many others got hacked,,, what says census won’t get hacked,, seriously wtf 🤬 I believe massive fine possible jail if not completed

• 
  XPRESS (@Xpress_24_7) reported 52 minutes ago

  🤖 𝗕𝗥𝗘𝗔𝗞𝗜𝗡𝗚: Telstra paid CEO Vicki Brady $6.8m after a nationwide outage; 20% of her bonus was docked but she still got close to $7m. Govt to introduce laws forcing tech platforms to pay for Aussie journalism: at least eight deals to acquit; 200% offset for SMEs. Source: The Guardian What happens next? Follow and like for more. #Aus #Tech #Media

• 
  Frank (@Frank93688755) reported 53 minutes ago

  This is the truth. Telstra engineers were required by HR to take a break. And that's why the system collapsed and emergency calls didn't go through. Whoodathunkit?

• 
  Nicole Kennedy 🇦🇺 (@kennedylnicole) reported 57 minutes ago

  @Telstra I have requested cancellation of a service on four separate occasions, yet continue to be charged multiple times. This morning your chat team again asked whether I wish to cancel it. Please confirm the cancellation under reference 130242508 and address my complaint 136862308 at your earliest convenience. Thank you.

• 
  Sue Davis (@Sueberry2) reported 58 minutes ago

  @blu_boys @Optus Go Aldi...roll over data, plans from $23, Telstra network, and currently double data on some plans for 6months if you start up before the end of September 2026.